Hello, Ive got access to a pair of GU-81 tubes for a bit of experimenting and I have a couple of design limits/safety parameters id like
to attemt to keep within. Ive built plenty of valve amps and have studied may topologies. This project is more along the lines of functional (barely) art,
-Desired power output is anything over 20w (yes 20w not 200w)
-Max voltages 800v (or even less) preferably using a tube rectifier
-OTL (Can/has this be done with a directly heated Pentode, (im thinking triode strapped operation, Single End)
-transformer input or interstage to isolate the amp from its input source.
-input, driver, and output stage topology.
-efficiency is irrelevent, it be usefull in winter.
-Im not taking this as a joke and will come up with something working but keeping the voltages low keeps the cost down remember a gu-81 is cheap compared even to a 2A3 or 300B
Has anybody gone down this path? With a bit of research theres a mode called inverted tube operation that got me thinking.. but honestly im looking at several of the 300b style OTL circuits out there to keep it simple. Id love some advice or comments from people playing with this tube. Ive been following many of the other posts and projects online so I know this is oddball and im sure it will gain some attention.
Cheers
Tim
to attemt to keep within. Ive built plenty of valve amps and have studied may topologies. This project is more along the lines of functional (barely) art,
-Desired power output is anything over 20w (yes 20w not 200w)
-Max voltages 800v (or even less) preferably using a tube rectifier
-OTL (Can/has this be done with a directly heated Pentode, (im thinking triode strapped operation, Single End)
-transformer input or interstage to isolate the amp from its input source.
-input, driver, and output stage topology.
-efficiency is irrelevent, it be usefull in winter.
-Im not taking this as a joke and will come up with something working but keeping the voltages low keeps the cost down remember a gu-81 is cheap compared even to a 2A3 or 300B
Has anybody gone down this path? With a bit of research theres a mode called inverted tube operation that got me thinking.. but honestly im looking at several of the 300b style OTL circuits out there to keep it simple. Id love some advice or comments from people playing with this tube. Ive been following many of the other posts and projects online so I know this is oddball and im sure it will gain some attention.
Cheers
Tim
Getting 20 Watt from an 8 Ohm loudspeaker requires 1.58 A of current. I don't see how a GU81 could provide that.
What is the internal resistance of a triode connected GU81?
What is the internal resistance of a triode connected GU81?
yes 1.5 amps is way ove, the pentode charts point towards 450ma at a fairly hefty plate voltage.
I havent found any triode charts yet, but im sure something will pop up with more research. Its early days with this idea and finding some good triode info will fairly quickly sink or float this OTL concept.
I havent found any triode charts yet, but im sure something will pop up with more research. Its early days with this idea and finding some good triode info will fairly quickly sink or float this OTL concept.
From these curves:
μ = dVa / dVg1 = 130 / 40 = 3.25
gm = dIa / dVg1 = 300 / 40 = 7.5 mA/V
Rp = μ / gm = 3.25 / 7.5 = 433 Ohm
As a cathode follower the output impedance would still be 1 / gm = 1 / 0.0075 = 133 Ohm.
μ = dVa / dVg1 = 130 / 40 = 3.25
gm = dIa / dVg1 = 300 / 40 = 7.5 mA/V
Rp = μ / gm = 3.25 / 7.5 = 433 Ohm
As a cathode follower the output impedance would still be 1 / gm = 1 / 0.0075 = 133 Ohm.
I agree that its clearly not a OTL tube, back to the OPT transformer it is
I just found this online load line calc.
https://www.vtadiy.com/loadline-calculators/loadline-calculator/
It has gu-81 as an option so im very impressed, if its close it does seem to indicate that a fair amount of usable power can be had between 600-800v in triode mode, with a load of 1200ohms at aprox160ma, ( i could be reading this wrong but 40w should be acheivable)
1200 ohm load is easy to mock up and test with, im sure a power transformer could drop in for testing (re gapp the transformer first)
I just found this online load line calc.
https://www.vtadiy.com/loadline-calculators/loadline-calculator/
It has gu-81 as an option so im very impressed, if its close it does seem to indicate that a fair amount of usable power can be had between 600-800v in triode mode, with a load of 1200ohms at aprox160ma, ( i could be reading this wrong but 40w should be acheivable)
1200 ohm load is easy to mock up and test with, im sure a power transformer could drop in for testing (re gapp the transformer first)
The dissipation of the two triode connected GU-81's is 110 Watt. How can that give an output power of 90 Watt?
Can C3 be an electrolytic capacitor? Doesn't its positive terminal swing more negative than its negative terminal half of the time with signal?
Can C3 be an electrolytic capacitor? Doesn't its positive terminal swing more negative than its negative terminal half of the time with signal?
110 Watts is Class A portion, the rest is B portion, peaking at 600mA or 400W some go to output, the rest is dissipated as heat. See attached for more info.
The output where C3 is connected should be < 100mV offset, even at full output. It may be a problem if the top bank is fused off and C3 connected to -ve rail. But there is how every otl (include solid state) will use to block DC.
The output where C3 is connected should be < 100mV offset, even at full output. It may be a problem if the top bank is fused off and C3 connected to -ve rail. But there is how every otl (include solid state) will use to block DC.
Attachments
90 Watt output power from two triodes dissipating 55 Watt each would mean an efficiency of 82 %. As far as I know, that is impossible.
Last edited:
This is Class AB amp, does not not need same dissipation as Class A, it only needs enough idle current to overcome cross over distortion. E.g.Tim Mellow OTL is Class AB, only dissipate (150*0.15) 22.5W, the output is 25W.
You sound like you're talking about a pure Class A amp.
You sound like you're talking about a pure Class A amp.
OK, thanks. My fault.
What would happen if the two GU-81 are not completely matched? Would the offset at the positive terminal of C3 stil be 100 mV positive with respect to ground?
What would happen if the two GU-81 are not completely matched? Would the offset at the positive terminal of C3 stil be 100 mV positive with respect to ground?
No, the offset would be more than 100mV, +or- depend on Mu or now the gain/loss of stages. It can be manually corrected by readjusted the bias until it's <100mV. Or can be auto-corrected when the entired amp has DC servo system built in (e.g again Tim' OTL) which sense the output terminal (looking for 0V or so). Having said that it does not mean that the mismatch issue is completely solved. One of the tube will start to clip earlier for example or offset detected when there is output power delivered to the load. That may be solved by negative feedback to correct gain of the stage.
Very interesting, I'm thinking totem pole is possibly the way to go. Looking at he curves I took a quick guess at an operating point within my limits. Even a single gu81 in class A appears to hit my power output target. 👍 Now onto the totem pole arrangement there's an opt transformer (which I do like for isolation and safety) 600ohm to 8 ohm opt, am I correct? I wonder if this would have an acceptable frequency response? Looking at totem pole the power out is great, correct me if I'm wrong but 1ch could make 100w+ and consume 750w
DC offset through the opt transformer. Am I right to think that in totem pole it will be tiny based on 100mv (this would make the opt fairly small )
In that configuration, you could capacitor couple the trafo and then not worry about offset or drift. You can even bridge two amplifiers.
If you‘re not opposed to using sand, you could implement a bias servo. It wouldn’t be handling the audio signal, just nulling out any offset.
If you‘re not opposed to using sand, you could implement a bias servo. It wouldn’t be handling the audio signal, just nulling out any offset.
Ireally like the idea of no DC in an opt transformer as it opens up lots of options. But +-700v gives the PSU a big potential between rails. I'm thinking psu layout safety. Going back to my thread starter I'm going to lean hard on building a single tube class A amp as it's clear I'll get over 20w. Having DC out of the opt is high on my list. Parafeed? or up run It as a pentode in ul and use big iron? I'd be hoping for very little distortion under 20w as honestly everything really relies on the first few watts.
Me too also scare of voltage over 500V. But not everyone agree why one should stick to +-700V. It can deliver more than 30W with +-300 rail, or a total 600V. With SE amp you need about the same, I think about 500V on g2. Remember if you want more power with this tube drive it into A2, that should double the output power. Attached is Class A, can mod to A2, but I don't agreed everything in the snapshot (not mine): 150W?,in pentode mode.
You can do para feed so you can use PP type OT, need to use suitable choke to drive the OT with sufficient current more if into A2.
Attachments
Last edited:
The (im)possibility of an electrolytic capacitor as C3 in the schematic in post #7 keeps nagging my brain.
For AC current to flow at the output, there must be voltage swing, or? If so, than the voltage at the positive terminal of C3 has to swing up and down, or? Since one side of the output terminals is grounded, the swinging will take place around ground potential, or? If all this is true, than the positive terminal of C3 swings both positive and negative with respect to ground, or?
If the above is true, how come that an electrolityc capacitor can be used as C3? I thought that the positive terminal of an electrolytic capacitor should always be positive with respect to the negative terminal, also under signal conditions.
So what am I missing. Am I wrong about the workings of the schematic? Or about what an electrolytic capacitor can withstand? Or am I wrong about both?
For AC current to flow at the output, there must be voltage swing, or? If so, than the voltage at the positive terminal of C3 has to swing up and down, or? Since one side of the output terminals is grounded, the swinging will take place around ground potential, or? If all this is true, than the positive terminal of C3 swings both positive and negative with respect to ground, or?
If the above is true, how come that an electrolityc capacitor can be used as C3? I thought that the positive terminal of an electrolytic capacitor should always be positive with respect to the negative terminal, also under signal conditions.
So what am I missing. Am I wrong about the workings of the schematic? Or about what an electrolytic capacitor can withstand? Or am I wrong about both?