• WARNING: Tube/Valve amplifiers use potentially LETHAL HIGH VOLTAGES.
    Building, troubleshooting and testing of these amplifiers should only be
    performed by someone who is thoroughly familiar with
    the safety precautions around high voltages.

KT88 and Hammond Transformer Schematics

Status
Not open for further replies.
KT88 start up

Gingertube
I still have not rounded up a scope but presntly waiting for the signal gen to arrive. I hope to have the equipment by this weekend or next week to take measurements. Any thought to my gain problems.
What measurements do you need for the verification of voltages at the input , 12ax7, 12at7, ky88 and the output tranny both side?
I would a sum using 1000 hz.
I believethe standard input voltage would be 2.5 to 3v max.

What would be the voltages at the above mentioned parts at full input ie the output tranny at 16 ohm 60 watt.
Secondly if the tranny and circuit is designed for 60 watts at 8ohm what will the output be for 16 ohm. Is it less as normally found in solid state designs.

Can't wait to burn more solder hope i don't burn the scope. Here we go learning the hard way to use a scope. Should have paid more attention in grade 12 high school, not that i could remember back 28 years.

Thanks Ian
 
KT88 start up questions

I have been testing the amp circuit and need some additonal information. I amp using a 12ax7 drive tube for the input and a phase inverter 12 t7 feeding a pair of kt88.
1. The wires running in the drive stage are severly affected be there relative location to each other. Can someone tell me which ones need to be isolated, twisted and or any other rules of thumb from your experiences.
The heater wires are twisted and located away from the signal to tubes.
The location of the wire affected the overall output . The x7 tube would only provide 2 volts peak to peak until i relocated them. I now get 3 volts prior to clipping.
2. Can someone also tell me the expected voltages on the drive stages and KT88 plates and grid. With 2 volts in Pk to Pk I get 22 volts pk to pk out at the 16 ohm tap on a passive resistive load. 3. The other channel only has an output of 20 volts which is a 10% difference. Is this normal?
4. It was calllculated that I should get 32 volt out at the load to reach maximum output or 60 watts. I'm only at 20 to 22volts at 2 volts in.
 
Gingertube

Hi Ian i see that your on line.
I sent information to you by email. I need help to finalize the amp circuitry. It stil is only producing about 20 watts output on 16 ohms or 13 volts rms (speaker terminals). I resent schematics and also scope readings with the T7 installed and removed (Kevins Request). I also changed the X7 plate resistor to 470k from 390k. I get about 19 volts out of the T7 tube feeding the KT88 stage. The voltage at the X7 is 26 with the T7 removed and drops to 19 with it in. The cathode follower is reducing the drive stage wich in itself is not enough to run the KT88's.

You mentioned the last time i might have to change the resistors from the transistor at the emitter and to ground. Kevin suggsted to hook the differential splitter into the T7. I sent you a schematic on this. Any thoughts

Stranded in Canada with 1/3 the output

Thanks Ian if you can help
 
Scott,
Yep here - was busy typing a response.
I think I gave you some bad advise earlier. Another look at your schematic shows +347V at the power supply to the front end diff amp with the tubes in and +536V with the tubes out.

I said to adjust the diff amp current source (via the transistor s base to 0V resistor) to get anode voltages of 270 to 300V.

That would be OK if the power supply was +536 with the tubes in - but it isn't. You should, instead, be adjusting the current source to get say +230V (+/- 20V).

The bias on the KT88 oututs is -68V. That tells you that for full power output you need to swing +/- 68Vpk or 136V pk to pk. With the rail at +347V and the anodes at +300V all you can swing is 47V pk and then, only by running the tubes into cuttoff.

With a rail of +347 and anode voltages of say +230V you scan theoretically swing +117V pk or 234V pk-pk.

So adjust the current source to get those anode voltages down and let us know how it goes.

Kevin's suggestion to hook the differential feed (the 2 x 1M2) resistors to the outut (cathodes) of the 12AT7 cathode followers rather than to the diffamp anodes is probably worth a try too. What it does is reduce the load on the diffamp and allows you to maximise gain and swing a few more volts.

If all this works OK then there is one other thing I would recommend and that is to modify again for more current in the 12AX7 diff amp - but that can wait untill this lot is working.

Cheers,
Ian
 
Hi Ian
My B+ is 550 volt with the tubes in and 600 with out.
THe voltage measurement atht e palte of the x7 is the reading taken with the 390k resistor in. I forgot to recheck that measurement. I think back when i first started up the unit i have a 100k plate reesistor and the voltage was down at the low 200 mark. The was even less output performance. When you had me change the resistor to 220 it improved and upon changing the emitted it was even better.
I checked back and at 220k plate with a 2k2 emitter resistor i had 230 volts at the plate.

Thanks
 
Hi Ian
I'm running the KT88 at 40mv or 40 ma approximately. If i increse this to 50 with the -68 volts reduce making the tubes easier to drive from the drive stage or does the neg voltage go up.
What is the max that the KT88 can take or is 40 max.
Thanks
 
Scott,
The KT88 is rated at 42 Watts - for decent lifetime run them at 35W. I see your main rail is 550V so 35/550 = 0.058 or 58mA

Run the tubes at 55 mA. This will reduce the required voltage swing to drive them a small amount BUT mostly it will reduce the output impedance of the amp by lowering the rp of the output tubes. That in turn means means you need less negative feedback for the same damping factor and less input voltage for the same output. Also there is general agreement that less negative feedback gives the amp a more "live feel".

Increase the tube currents now but leave the feedback adjustment until you have everything else done.

Cheers,
Ian
 
Thanks Ian
I will play with the bias seting when i get back. I am also waiting to see what Kevin and Dante come up with. I also looked at the dnaco circuit to see the differences. The dyncao uses a semicinductor also i belive for the splitter after the second tube. It uses one and one half tube for its gain, Can anyone telll me what the voltage would be from the second tube to the kt88. If i dont get results i might try the dynaco drive stage and leave the power supply and KT88 biasing circuit as is.

The dynaco uses a LM334 semiconductor. Is there another NTE number to crossreference it

Thanks
 
KT88 start up

Hi Ian back now from Quebec. I have not recieved any further information from Kevin or Dante . I will have to play with the bias setting an see what happens. I will increase the setting from 40 mv to 55 mv and see if there is more output. I guess i should check what the negative voltage is at this point to see what is required to drive the KT88 to full output.

Did you get the scope read outs

Thanks
 
KT88 burnout

Gingertube

Hi Ian didn't have much success
I tried increasing the bias voltage from 40 ma to 55. This only resulted about 2 more volts on output (13 volts rms at 16 ohms) and also a runnaway tube within 30 seconds . Blow the main power fuse and scorched a kt88. I have to pick up one new pair tomorrow. The kt 88 still work but starts to clip at 3/4 input. The metal plates inside are darker and even the painted label on the glass has discoloured.

I'll see if there are any sugestions from Kevin. He was looking at relocating the splitter to the follower t7plates instead of the x7.

I think i sent you that circuit for the 1m2 relocation

I looked at the dynaco MArk III circuit, the posiden board application. They use half of the x7 for the first gain stag and afull 12a7 for the second. They also use a LM334 ( application specific semi conductor from National) for the splitter. My existing power supply is higher 550 B plus and the dynaco calls for 480. The V plus is presently at 372 volt and the dynco calls for 430 volts. I'm sure i could acheive the voltages with resistor changes if required

Can i use the front end of the dynaco circuit and leave the power tubes, and biasing arrangement still the same including all the transformer conections. I am proposing building the drive stage only.
Would you review the proposed circuit incomparison to the existing for me. I can send you both schematics

I have the parts to build the circuit but preffer to have someone review.

I will use the same power supply solid state and modifiy the voltages if nesecary.


Thanks Ian


PS
Can anyone tell me what the drive voltages are on A dynaco Mark III, from the 12a7 to the KT88 or 6550 before clipping.
 
Scott,
You could certainly use the Poseidon Driver Board. There are also other after market driver boards which would suit and might even be superior (CAE for example).

Before embarking on a major rebuild why not wait to see what Kevin comes up with - its his design after all.

It seems to me that the existing amp is not that far off. The only real reason for a major surgery at this point is if you really do not have enough gain. This (lack of gain) is what Kevin is probably trying to address by the suggestion of using the 12AT7 as a splitter/driver and the 12AX7 as just a preamp stage. No reason why that can't work well. Have a look through POP. One alternate topology is to keep the 12AX7 diffamp input and convert the 12AT7 from cathode followers to a differential driver. I'm fairly sure that there was an example circuit in the section dedicated to Toroidal Output Transformers, look for any design which was intended for PAT4002 Output Transformer.

Changing to a differential driver may require you to change from the 12AT7 to something "beefier" say a 12AU7 of better still an ECC99 because you have to drive those 100K grid resistors on the KT88 - don't be tempted to increase those, particularly in view of the fact that you had a tube run-away, any increase in those resitors makes a ru-away more likely.

Cheers,
Ian
 
Scott,
Look here:
http://www.curcioaudio.com/

I actually have a pair of the MKIII Premium Mod Boards. One of them is working beautifully in a prototype 120W Ultralinear with 4 off KT88 driving a Hammond 1650T. The other is sitting on the shelf waiting for me to finalise the design and then build the second monoblock. I also have VDV2100 and VDV2100 CFB/H Plitron Toroidal Output Tannies and haven't made up my mind if to stick to the Hammond 1650T or put the expensive ones in. This is a project that has been on the "back burner" for a while.

The driver board is basically the same as your 12AX7 front end BUT instead of a single 12AX7 triode on each side of the diff amp it uses cascode connected 6DJ8 on each side of the diff amp. The board also has a dedicated high voltage regulator on it. The cascode 6DJ8 (6922, ECC88) gives more gain than a single 12AX7 triode.

Cheers,
Ian
 
KT88 start up

Thanks Ian

I think you are saying you use a different tube inplace of the x7.
I don't have a clue what the regulator is. I need to get this fix soon to start the 120 watt amp. I"d like to finish them before spring settles in.

If you could send me a sketch of the circuit i probably could understand better.

I don't want to persue the posiden circuit and really what to fix the existng.

Thanks Ian

I think it's time for pair if these

www.lenardaudio.com/05_opal.html

Cutting out for the evening Bye
 
KT88 Startup Burnout

Gingertube

High Ian picked up brand new set of JJ KT88 today. I started with 20mv 0n the bias setting which should be 20 ma. I previously was running the last set at 40mv with out a problem. I could put 1 volt rms in and would get 13 volts rms out on a 16 oh load.

So i installed the new tube and set the bias to 20mv. I then applied 1volt @ 1000 hz in and got 10 volts rms out at the load ( resistive). I rechecked the bias seting while running the input and load. The bias seting jumped up to 30mv yet returned to 20 when the input was reduced to zero.

I then set the bias to 30mv with no input . I applied 1 volt rms in and got 11.5 volt out and the bias jumped to 42mv.

My first question is this normal for the bais to raise as the signal input increases?

Secondly by raising my bias i only achieve 1.5 vilot more outout. I the calcuatioins are right i need 32 volts at the 16 ohmload to achieve full output. THerefore i only about 3 watts by increasing the bias from 20 to 30 mv.

I cant put anymore than 1 volt rms in before the t7 output starts to clip. I am only getting 12 volt rms out. Im barely reaching half output.

Anyway the interesting thing is it sounds beautiful compared to my bryston and parasound even as defective as it is. I can set the input higher with the electronic 4 way crossover. I wish you could here. I need to finish the trouble shooting to start the 3 other amps. I have no doubt my speakers will match or better the peromance of the last link i posted

Thanks Ian

PS if there is anyone else in the Toronto London area that could help please responde
 
Scott,

Busy typing the bottom bit of this when the email alert went off to say you'd posted.

Edit: I dont think its the T7 starting to clip. Its more likely to be the input stage - you are running out of compliance voltage in the cathodes of the 12AX7 to run the current source. You may need to return the bottom of the current source to a negative rail or ditch the current source and use a resistor as per the example from POP sited below.

See also Fig 5-24 on page 5-30. This would be easier to try first. Just genearte the negative rail off the bias supply and use a tail resistor in place of the current source.l

Always measure the bias with no signal. You will get strange readings (as you noted) when a signal is present.

If you decide to convert the 12AT7 from cathode followers to a diffamp driver (for more gain) look at POP page 5-28 Fig 5-22


Note that you need to add a R and a C to get another decoupled positive supply and an R and a C to get a negative decoupled supply from the bias power supply for the cathode resistor to connect to.
Watch the voltage rating of the caps. At power on, before the tubes heat up and start conducting the voltages will be very high. You may need to series caps like you did for the main HT caps.


Cheers,
Ian
 
kt88

Thanks Ian is good to know the fluctuation is okay. I'm sitting at home listening to these new tubes. I probably only use 1 watt at most.

I will try and call Kevin to see the next step.

I am going to make 2 more stereo chassis to start. My concern is that if i would have to add another drive tube into the chassi. I only hope that i can make due with 2 drive stag tubes. I have enough room inside to do any other rmods. Your thoughts would be appreciated.

I have all the components to make the posiden front end and london power, power supply and biasing system. I have a pair of 12au7 tubes also.

I won't start the 120 monoblock chassis untill the front end is solved.

I have all the trannies only need a grand for the tubes and socket. Thats a short drive and that will be solved.

I figure when i have all the amps running the buiding will self heat. It's about -10 out. I'll have to get ac for the summer.


Thanks
 
Thanks i got my gas bill $280 for only months heat in January wich was a record for above normal temp. This month is colder so it should hit $350.

Move to Canada and keep the utility companies rich.

Bed time

PS will i get buy with 2 drive tube?

Thanks
 
Status
Not open for further replies.