Hi, this is my first post and am looking for advice/resources on using 807's and 1625 tubes, plus am posting a cool 1954 schematic below.
I'm still searching a bit (not enough time in one day, lol) and found
so far these amps based on the 807 or 1625:
A low power amp based on 807's:
http://www.tubes.mynetcologne.de/roehren/807pp/807pp_e.html
A single-ended one here with not to many details (the main site went offline and this is just a copy):
http://www.electra-print.com/images/1625schem2.gif
What I'm trying to do, basically, is come up with a low cost, higher-power amp using 1625's, which are basically the equivalent of the 807, except having a 12.6v filament and offbeat 7-pin socket.
The 1625's are war surplus in big supply, and are ony about $5 just about everywhere for NOS, and can do up to 100w with just 1 tube, they say.
A 2-tube push-pull would probably be my first choice, not something huge like the Dogzilla on Fred Nachbauer's site:
http://www.dogstar.dantimax.dk/tubestuf/dzindex.htm
Not this by any means, nor a too time-consuming project.
I'm not an engineer, but was into a bit of ham stuff and one point, and basically know a few musicians that could appreciate a homebrew higher-end rig.
I found this in in 1954 old Amateur Radio handbook (images are large but I wanted to get the detail), I hope some of you can appreciate this schematic and writeup. I put the 2 images on my own site for now:
http://viewoftheblue.com/misc/807_1.png
http://viewoftheblue.com/misc/807_2.png
As for this one, it says "Push-Pull 807 Modulator & Speech Amplifier", but can this, as shown, be used as a single standalone unit unit going all the way from the mic to a speaker output transformer at the other end.? (they seem to be showing it as going to another unmentioned stage).
It says it can do 100 watts with enough voltage, and only has 3 other smaller octal tubes and shows 2 transformers for the entire unit (not including the power supply, I believe).
I haven't researched the T2 unit yet, but it seems that it has the same turns ratio for the primary & secondary, except for a center tap, unless they mean 2:1 and then a center tap also (it's not clear)
I guess a fairly hefty output transformer would be needed as well.
I've seen some of the items at Hammond as well as Angela, but would appreciate more/better sources,etc..
One the second page it does show a power supply schematic with an adjustable bias control, but doesn't really give the specs on the transformer or voltage. I'm seeking advice on a higher-end setup to take more advantage of the tubes and would appreciate learning what secondary voltage (center tapped) would be best and how many milliamps it would need to be rated. If price becomes an issue, we could underdrive it, use smaller power & output transformers and shoot for, say 35-50 watts. Just need some advice here, and especially if anyone knows of a similar rig on the net with a schematic & more specific parts/construction inof listed.
The second link (electra-print com) looks like it runs in ultralinear but only puts out 16.5 watts if I'm seeing it correctly, with one 1625. My big question with this circuit is why he is providing DC to the filament rather than AC, has it something to do with the preventing parasitic oscillations (as the chokes take of in the old 1954 project)?
So basically any advice on finding reliable (hopefully free) schematics on similar projects (or even just more construction info if this one will do nicely) for this tube, and transformer advice, etc.
Hopefully just using 1 power transformer (except perhaps an extra to cover an extra un-provided-for filament voltage)
I notice on angela.com they seem to have a good but not exhaustive slection of high-end transformers,
Allied Electronics has one big one but it's only rated at:
750VCT @ 150 mA , and 5v @ 3A, with a 6.3 CT @ 5A but it's only $43.00
Anyway, hope you enjoy the schematic, and thanks in advance for any advice.
Frank
I'm still searching a bit (not enough time in one day, lol) and found
so far these amps based on the 807 or 1625:
A low power amp based on 807's:
http://www.tubes.mynetcologne.de/roehren/807pp/807pp_e.html
A single-ended one here with not to many details (the main site went offline and this is just a copy):
http://www.electra-print.com/images/1625schem2.gif
What I'm trying to do, basically, is come up with a low cost, higher-power amp using 1625's, which are basically the equivalent of the 807, except having a 12.6v filament and offbeat 7-pin socket.
The 1625's are war surplus in big supply, and are ony about $5 just about everywhere for NOS, and can do up to 100w with just 1 tube, they say.
A 2-tube push-pull would probably be my first choice, not something huge like the Dogzilla on Fred Nachbauer's site:
http://www.dogstar.dantimax.dk/tubestuf/dzindex.htm
Not this by any means, nor a too time-consuming project.
I'm not an engineer, but was into a bit of ham stuff and one point, and basically know a few musicians that could appreciate a homebrew higher-end rig.
I found this in in 1954 old Amateur Radio handbook (images are large but I wanted to get the detail), I hope some of you can appreciate this schematic and writeup. I put the 2 images on my own site for now:
http://viewoftheblue.com/misc/807_1.png
http://viewoftheblue.com/misc/807_2.png
As for this one, it says "Push-Pull 807 Modulator & Speech Amplifier", but can this, as shown, be used as a single standalone unit unit going all the way from the mic to a speaker output transformer at the other end.? (they seem to be showing it as going to another unmentioned stage).
It says it can do 100 watts with enough voltage, and only has 3 other smaller octal tubes and shows 2 transformers for the entire unit (not including the power supply, I believe).
I haven't researched the T2 unit yet, but it seems that it has the same turns ratio for the primary & secondary, except for a center tap, unless they mean 2:1 and then a center tap also (it's not clear)
I guess a fairly hefty output transformer would be needed as well.
I've seen some of the items at Hammond as well as Angela, but would appreciate more/better sources,etc..
One the second page it does show a power supply schematic with an adjustable bias control, but doesn't really give the specs on the transformer or voltage. I'm seeking advice on a higher-end setup to take more advantage of the tubes and would appreciate learning what secondary voltage (center tapped) would be best and how many milliamps it would need to be rated. If price becomes an issue, we could underdrive it, use smaller power & output transformers and shoot for, say 35-50 watts. Just need some advice here, and especially if anyone knows of a similar rig on the net with a schematic & more specific parts/construction inof listed.
The second link (electra-print com) looks like it runs in ultralinear but only puts out 16.5 watts if I'm seeing it correctly, with one 1625. My big question with this circuit is why he is providing DC to the filament rather than AC, has it something to do with the preventing parasitic oscillations (as the chokes take of in the old 1954 project)?
So basically any advice on finding reliable (hopefully free) schematics on similar projects (or even just more construction info if this one will do nicely) for this tube, and transformer advice, etc.
Hopefully just using 1 power transformer (except perhaps an extra to cover an extra un-provided-for filament voltage)
I notice on angela.com they seem to have a good but not exhaustive slection of high-end transformers,
Allied Electronics has one big one but it's only rated at:
750VCT @ 150 mA , and 5v @ 3A, with a 6.3 CT @ 5A but it's only $43.00
Anyway, hope you enjoy the schematic, and thanks in advance for any advice.
Frank
If you're after high end, then you can forget about all those modulator shemos from the ARRL Handbook. Useage as AM modulators finds hiFi neither deireable, nor legal. "Circuit constants have been selected so that the overall fequency response is sufficiently flat in the normal range of voice frequencies, but drops off above 3000 cycles and below 150 cycles." That translates to:
An externally hosted image should be here but it was not working when we last tested it.
as to how that will sound. Also, his claim that he's getting "approximately" 200W is either very approximate or that sucker is eating 807s like mad. The RCA spec sheet for the 807 promises a maximum of 120W for deep Class AB2, again, not a Q-Point with audio excellence in mind, but rather lotsawatts for modulator service.That was obviously the main design criterion: max power and efficiency with little to no consideration given for the sonics. Also, the design has way more front end voltage amplification than you'd ever need for any reasonable HiFi amp.
According to the STC Application Report, you can operate a bit more conservatively:
Vpp: 600Vdc
Vsgsg: 300Vdc
Vgk: -30Vdc
Ipq: 60mA
Isgq: 5.0mA
R(L): 6.4K (p-2-p)
Pi: 100mW (total)
Po: 80W
THD: 3.5%
Not to bad at all. (For my 807 project, I opted for the 1.8%THD, 26.5W, Class AB1 operation) Add some local NFB to reduce the effective Rp of the 807s, and some gNFB to clean up any remaining pentode nastiness, and it could sound very good indeed. I would also consider active voltage regulation for the 807 screens.
That'll sound a helluvalot better than those AM modulator schemos.
An externally hosted image should be here but it was not working when we last tested it.
OK, after all this time I did some research and find that there are a few good 807 designs, aside the old Williamsons.
For background, the one which most people try is the old (fairly straight Williamson) Radiotron A515 (I'm talking hi-fi rather than guitar amps in this case):
http://www.retrovox.com.au/A515.pdf
Here's a site that mentions a lot of good tweaks patch up some flaws in these original circuits:
http://home.att.net/~chimeraone/tubeampmodswilliamson.html
I ran across the site with the "Heffa" amps and it's a public domain design and has an extensive how-to.
The Heffa amps are designed to be economical to build and run in pure class A (see the writeup), and very clean sounding, but low power (only 10 watts or so per channel for push-pull 807's). This does sound like a very clean amp and worthwhile project.
The power supply is custom and mostly of European interest, so we here in the US would have to hack one together, which wouldn't be difficult.
Here's the main site:
http://heffa.ath.cx/heffa/
Here's the schematic:
http://heffa.ath.cx/heffa/heffa/heffaact.pdf
The schematic shows just one channel, where most of the people build one with two channels for stereo, though some do build a monoblock.
In their technical pdf, they mention as similar circuit they highly respect, the DIY project by Darrell Whitfield using 4 807's to make a parallel push-pull circuit at 24 watts, which is on the Triode electronics site:
http://www.triodeel.com/tube0521.gif
Now, I'm thinking that a combination of the two would be a good way to go for a personal project. Keeping in mind that I'm not an engineer, I came up with a combined schematic. using the 12SN7's as in the Heffas for the pre-amp stages and then took Whitfield's final stage for the rest of it. Please take a gander at this and see if you think the components are in range, and I didn't screw up on the main concept:
(Not sure if my voltages are exactly right, and I also added a volume /tone control of 2 audio pots, which I saw in some guitar amp circuit and am not sure if this will work, the volume and tone are both 100Meg Audio tapers, according to that, and going to the first grid of the second 12SN7).
Please excuse my drawing, I only put about an hour into it, then had to scan, crop and resize it.
So if this prelimarily seems fairly sound, I can think about the power supply, output tranny, etc. And actually I'll be using 1625's instaed of 807's
according to the Heffa tech info, the circuit is self-balancing too, for tubes that are off a bit, and correct the performance automatically.
For background, the one which most people try is the old (fairly straight Williamson) Radiotron A515 (I'm talking hi-fi rather than guitar amps in this case):
http://www.retrovox.com.au/A515.pdf
Here's a site that mentions a lot of good tweaks patch up some flaws in these original circuits:
http://home.att.net/~chimeraone/tubeampmodswilliamson.html
I ran across the site with the "Heffa" amps and it's a public domain design and has an extensive how-to.
The Heffa amps are designed to be economical to build and run in pure class A (see the writeup), and very clean sounding, but low power (only 10 watts or so per channel for push-pull 807's). This does sound like a very clean amp and worthwhile project.
The power supply is custom and mostly of European interest, so we here in the US would have to hack one together, which wouldn't be difficult.
Here's the main site:
http://heffa.ath.cx/heffa/
Here's the schematic:
http://heffa.ath.cx/heffa/heffa/heffaact.pdf
The schematic shows just one channel, where most of the people build one with two channels for stereo, though some do build a monoblock.
In their technical pdf, they mention as similar circuit they highly respect, the DIY project by Darrell Whitfield using 4 807's to make a parallel push-pull circuit at 24 watts, which is on the Triode electronics site:
http://www.triodeel.com/tube0521.gif
Now, I'm thinking that a combination of the two would be a good way to go for a personal project. Keeping in mind that I'm not an engineer, I came up with a combined schematic. using the 12SN7's as in the Heffas for the pre-amp stages and then took Whitfield's final stage for the rest of it. Please take a gander at this and see if you think the components are in range, and I didn't screw up on the main concept:
(Not sure if my voltages are exactly right, and I also added a volume /tone control of 2 audio pots, which I saw in some guitar amp circuit and am not sure if this will work, the volume and tone are both 100Meg Audio tapers, according to that, and going to the first grid of the second 12SN7).
Please excuse my drawing, I only put about an hour into it, then had to scan, crop and resize it.
So if this prelimarily seems fairly sound, I can think about the power supply, output tranny, etc. And actually I'll be using 1625's instaed of 807's
according to the Heffa tech info, the circuit is self-balancing too, for tubes that are off a bit, and correct the performance automatically.
I sort of wanted to be thorough in my research, etc, in the previous post with all my links, etc., but ignoring these, the main thrust of my question is basically if the last item (the schematic which I drew up and which appears as a pic at the end of the post) is do-able, and won't blow anything up (as is). Basically I combined 2 designs and am not sure if the coupling or values are functionally correct.
I have not tried the 807 but, by all accounts, it is excellent if used well. Bearing in mind its limited screen voltage, it seems best suited to either beam tetrode operation or UL operation using a separate OPT winding for the screens.
I have read, however, that there are some good and some not so good examples of the 807 on the market, maybe because of the difficulty of building it with enough precision. This article explains a bit about its history and the problems of finding good ones during WW II.
I have read, however, that there are some good and some not so good examples of the 807 on the market, maybe because of the difficulty of building it with enough precision. This article explains a bit about its history and the problems of finding good ones during WW II.
I'm going to put any projects using these tubes on hold, because I never could find good design I liked, and I don't really think at this point, wasting a lot of heat on a class A design woud be the way to go.
The big classic using the 807's was the 1954 Standel 25L15 guitar amp using 807's, and it supposed to sound sweet almost hi-fi in quality, but in all my web search travels it's been one of the best kept secrets, and no one has really let loose a good schematic on it.
They are selling megabuck repros of this amp, but the schematics are kept top secret.
The big classic using the 807's was the 1954 Standel 25L15 guitar amp using 807's, and it supposed to sound sweet almost hi-fi in quality, but in all my web search travels it's been one of the best kept secrets, and no one has really let loose a good schematic on it.
They are selling megabuck repros of this amp, but the schematics are kept top secret.
Hello Frank,
From my own search (got some 1625's to put to work) I've found these. As you can see I'm more attracted to the recent audiophile designs.
http://www.triodedick.com/topcap_807pp/topcap_schema.gif
http://www.triodedick.com/topcap_807pp/topcapv_schema_voeding.gif
http://www.dhtrob.com/schemas/afbeeldingen/807pp_verst.GIF
http://www.dhtrob.com/schemas/afbeeldingen/807pp_voeding.GIF
http://tinpan.fortunecity.com/saints/668/primer/
Simon
Time to Blow My Own Horn
EEEEWWWW!!!!
http://www.triodedick.com/topcap_807pp/topcap_schema.gif:
Problems:
http://www.dhtrob.com/schemas/afbeeldingen/807pp_verst.GIF
Problems:
In both cases, these designs use cathode bias which gives lots more THD than does fixed bias. (See: STC 807 Application Report) In neither case is there adequate screen voltage regulation, not even including a simple VR tube regulator.
If you want a really good design, then use mine:
Main Schemo
Screen Regulator
Power Supply
Here, I have adequate open loop gain, so that sensitivity will not be a problem under feedback. The phase splitter used here is the vastly superior LTP splitter with a solid state CCS for good AC balance, both in the fundamental and inevitable harmonics. The 807s operate under fixed bias, and the grids have their own drivers so that neither blocking distortion or slew rate limitations become a problem. The front end has a very generous amount of "headroom". Indeed, the OPTs show signs of core saturation before there is any actual clipping. The drivers can push the 807s into Class AB2 transparantly, so that this design actually 30W of output at the onset of core saturation, even though it's rating is 26.5W of Class AB1 power. The screen grid voltage is regulated by an active regulator since screen voltage stability is a key to low distortion operation. Local NFB is included to reduce the effective r(p) of the 807, so that the xfmr primary inductance becomes a larger percentage of r(p). This helps improve bass and serves to get rid of that pentode nastiness. A moderate amount of gNFB further improves bass by reducing Zo, and cleans up what nastiness remains.
This 807 design beats the daylights out of anything I've seen elsewhere on the 'Net. If you're gonna use 'em, then use 'em right and get the most out of them. After all, 807s have some of the best THD figures of any power pentode.
Klimon said:
EEEEWWWW!!!!
An externally hosted image should be here but it was not working when we last tested it.
http://www.triodedick.com/topcap_807pp/topcap_schema.gif:
Problems:
- Very inferior phase splitter that is not balanced.
- 6SL7s can't properly drive the grids, especially at higher frequencies.
- Inadequate open loop gain.
http://www.dhtrob.com/schemas/afbeeldingen/807pp_verst.GIF
Problems:
- Slight improvement in the phase splitter, but still has problem of unbalanced harmonic distortion.
- Way too much gNFB. You don't need 20db(v) of gNFB for any reasonably well done 807 design. That'll sound very solid statey (and I don't mean good solid state either) and you will likely have stability issues. Given the choice of phase splitter (adequate for the simplest projects and/or where fidelity is not a concern) I'd say that this is an inadequate open loop design and he's trying to compensate with massive gNFB. This is not the way to employ gNFB.
- Way too little open loop gain, and likely sensitivity problems.
In both cases, these designs use cathode bias which gives lots more THD than does fixed bias. (See: STC 807 Application Report) In neither case is there adequate screen voltage regulation, not even including a simple VR tube regulator.
If you want a really good design, then use mine:
Main Schemo
Screen Regulator
Power Supply
Here, I have adequate open loop gain, so that sensitivity will not be a problem under feedback. The phase splitter used here is the vastly superior LTP splitter with a solid state CCS for good AC balance, both in the fundamental and inevitable harmonics. The 807s operate under fixed bias, and the grids have their own drivers so that neither blocking distortion or slew rate limitations become a problem. The front end has a very generous amount of "headroom". Indeed, the OPTs show signs of core saturation before there is any actual clipping. The drivers can push the 807s into Class AB2 transparantly, so that this design actually 30W of output at the onset of core saturation, even though it's rating is 26.5W of Class AB1 power. The screen grid voltage is regulated by an active regulator since screen voltage stability is a key to low distortion operation. Local NFB is included to reduce the effective r(p) of the 807, so that the xfmr primary inductance becomes a larger percentage of r(p). This helps improve bass and serves to get rid of that pentode nastiness. A moderate amount of gNFB further improves bass by reducing Zo, and cleans up what nastiness remains.
This 807 design beats the daylights out of anything I've seen elsewhere on the 'Net. If you're gonna use 'em, then use 'em right and get the most out of them. After all, 807s have some of the best THD figures of any power pentode.
Your design looks great and seems to have had a lot of thought put into it. I printed it out for future reference. Not being an engineer though, I would need the complete component specs, ratings, etc. before I could try it. Hopefully, in a few months I'll have more time and if you still stand by this design I'd be willing to get with you to try it, perhaps offer you a bit for your time in drawing up the exact specs.
Thanks MP, and I see you do post quite a bit here. I'm fairly new getting back into this hobby, after 30 years of lapse when I used to be into ham radio & built stuff back then, so I'm just doing a few simpler projects right now before getting into something intensely exacting and fine tuned such as this.
I didn't trust those other guys' plans either, though I had come across most of these designs that are available on the web. I see that the main critereon for this tube is, as you say, a certain degree of regulation.
Thanks MP, and I see you do post quite a bit here. I'm fairly new getting back into this hobby, after 30 years of lapse when I used to be into ham radio & built stuff back then, so I'm just doing a few simpler projects right now before getting into something intensely exacting and fine tuned such as this.
I didn't trust those other guys' plans either, though I had come across most of these designs that are available on the web. I see that the main critereon for this tube is, as you say, a certain degree of regulation.
I already have all the component specs anddesign details, but that'll have to be attached to an e-mail. You can get ahold of me at: ar_phox (at) inbox (dot) com.
Miles,
It occurs to me that if you take the local NFB loops back to the plates of the 6SL7, i.e. before C4 and C5, you wouldn't need C9 and C10 and you should also get a superior bass response, because C4 and C5 would not be looking into such a low impedance.
It occurs to me that if you take the local NFB loops back to the plates of the 6SL7, i.e. before C4 and C5, you wouldn't need C9 and C10 and you should also get a superior bass response, because C4 and C5 would not be looking into such a low impedance.
I already thought of that, but opted to keep the AC and DC paths separate, so that the recommended 6.95db(v) of local feedback (as recommended by O. Schade) would remain manageable. Based on what I'm hearing as well as measuring, he certainly knew what he was talking about there. So far as the impedance goes, C4-5 were selected to take that into consideration (0.22uF). The frequencies at the -3.0db points are: f(l)= 7.0Hz; f(h)= 37.5KHz. The gNFB remains effective at bass frequencies for decent damping, and the bass doesn't lack for "authority" without the underdamped "sloppiness" you hear so much about with VT designs. The phase linearity at LF is pretty good (as advertised by Hammond) and there's no tendency for instability at either end, even when too much gNFB was originally applied (sounded very solid statey).
Miles,
I see you've thought of it. BTW, what DF do you get? I'm inclined to think, by what I've read, that a DF > 10 is probably unnecessary unless the speakers are very low efficiency and rely on a lot of electrical damping.
I see you've thought of it. BTW, what DF do you get? I'm inclined to think, by what I've read, that a DF > 10 is probably unnecessary unless the speakers are very low efficiency and rely on a lot of electrical damping.
This is just the thread I have been looking for. I have quite a few 1625's that I am plannng to put to good use and I have been looking for a well designed amp to build. Nice job Miles! I am not a electrical engineer but I do know enough that a lot of the older designs on the net are not really of the audiophile design. I found this one that I have been reviewing http://www.triodeel.com/807-37.gif and your thoughts would be highly appreciated! I would be omitting the pre-amp section though. What I like is the output rating, which is what I am looking for, somewhere in the neighborhood of 75 to 100 watts. Cheers
ray_moth said:
The calculated DF is a little over 4.0. Using Bose speeks, which don't seem to be terribly inefficient, this works quite well. With just the local NFB operational, the DF was just a bit over 1.0, and though a great improvement over running open loop, obviously not good enough due to woofer resonance. As it works now, the bass packs plenty of "authority" without any sloppiness.
Toobstheway said:
Given the enormous front end gain (pentode + two triode stages) and the 800R load impedance, this looks like a design for a modulator. As it stands now, it would be wholly inadequate for any sort of HiFi audio. The phase splitter an especially bad design and there is no provision for balancing the currents of the finals. That suggests quick 'n' dirty, which would be consistant for an AM modulator where good sonics aren't necessary, and wouldn't be legal on the ham bands since the AM BW is limited to 6.0KHz. It is unusual to include gNFB in a modulator design, but not totally unheard of either.
You can operate 807s as Class AB2 amps to get ~80W, and do so with reasonable THDs and good sonics, but this design ain't it.
That's whay I've been so cautious too, since my first post on this thread. These tubes were very popular among hams and makers of cheap voice (PA's, modulators for AM transmitters, etc.) amps, but not too many good designs for quality music purposes. And from what I've read they only do well with decent regulation.
Miles, I sent you an email to hopefully look into your design further.
Those class A designs above were nice, but run hot with low power,
somewhere there will be a winning design of the best compromise of cross-spectrum sound quality as well as a decent amount of power using this tube.
Miles, I sent you an email to hopefully look into your design further.
Those class A designs above were nice, but run hot with low power,
somewhere there will be a winning design of the best compromise of cross-spectrum sound quality as well as a decent amount of power using this tube.
frank754 said:
If you want "good designs for quality music purposes", look for 6L6-based designs. The 807/1625 are identical to the 6L6, except that these can run at higher voltages due to the plate connection at the top, rather than at the base. sure, hams used them because they were already in their xmtrs.
I don't see any reason why these can't be run as AB2's for that. The STC (elsewhere in the thread) gives the Q-Point, Rl, and THD estimates for some 80W of Class AB2 power, and at THDs that are comparable to the THDs of other audio-type power finals. I went with some 30W of Class AB1 because it's somewhat better, and I really don't need that much.