12-14 tube DIY M60 OTL (filament transformer)
Anyone know this company? ------->signaltransformer.com
I got a lots of hit on here while doing a search.
Those that are look for filament transformer for their DIY M60 OTL Project
I'm looking at ( MPI-200-12 ) for 12-14 tube DIY M60 OTL.
Electrostatic Shielded also
Your Options Appreciated
Thanks!
dwhitf
Power Supply - Multi-purpose international transformer
Description:
The MPI series transformers feature higher volumetric efficiency for improved performance compared with the conventional
50/60Hz transformers. They also incorporate international safety features making them ideal for worldwide applications. The MPI series can easily
be modified to your specific application
• General Specifications
• Power - 200 VA to 900 VA
• Dielectric Strength - 4000 Vrms Hipot
• Primaries - Dual / tapped primaries 100 V, 115 V, 200 V, 215 V, 230 V - 50/60Hz
• Dual Secondaries - Series or parallel
• Electrostatic Shield - Solid copper full width foil
• Touch-Safe terminals (IP20 type) offer a screw/binding clamp for hard wiring and a 3/16” &
1/4” Fast-On connection.
• Leakage Current @ <100 micro Amps
• Insulation System - Class F, 155* C
• Agency Certifications
• UL 1446, File # E66312
• UL recognized to UL 506 / UL 5085-2, File # E63829
• CSA certified to C22.2 No. 66.1, File # 221070
• Designed to meet VDE 0805 and VDE 0550
• TUV Rheinland certified to IEC EN61558-2-4 and EN61558-2-6, License # R7204249
http://www.signaltransformer.com/sites/all/pdf/MPI.pdf
Secondary
Part Number Series Parallel
MPI-200-12 12 VCT @ 16.7 A 6 V @ 33.3 A
MPI-250-12 12 VCT @ 20.8 A 6 V @ 41.7 A
MPI-300-12 12 VCT @ 25.0 A 6 V @ 50.0 A
Anyone know this company? ------->signaltransformer.com
I got a lots of hit on here while doing a search.
Those that are look for filament transformer for their DIY M60 OTL Project
I'm looking at ( MPI-200-12 ) for 12-14 tube DIY M60 OTL.
Electrostatic Shielded also
Your Options Appreciated
Thanks!
dwhitf
Power Supply - Multi-purpose international transformer
Description:
The MPI series transformers feature higher volumetric efficiency for improved performance compared with the conventional
50/60Hz transformers. They also incorporate international safety features making them ideal for worldwide applications. The MPI series can easily
be modified to your specific application
• General Specifications
• Power - 200 VA to 900 VA
• Dielectric Strength - 4000 Vrms Hipot
• Primaries - Dual / tapped primaries 100 V, 115 V, 200 V, 215 V, 230 V - 50/60Hz
• Dual Secondaries - Series or parallel
• Electrostatic Shield - Solid copper full width foil
• Touch-Safe terminals (IP20 type) offer a screw/binding clamp for hard wiring and a 3/16” &
1/4” Fast-On connection.
• Leakage Current @ <100 micro Amps
• Insulation System - Class F, 155* C
• Agency Certifications
• UL 1446, File # E66312
• UL recognized to UL 506 / UL 5085-2, File # E63829
• CSA certified to C22.2 No. 66.1, File # 221070
• Designed to meet VDE 0805 and VDE 0550
• TUV Rheinland certified to IEC EN61558-2-4 and EN61558-2-6, License # R7204249
http://www.signaltransformer.com/sites/all/pdf/MPI.pdf
Secondary
Part Number Series Parallel
MPI-200-12 12 VCT @ 16.7 A 6 V @ 33.3 A
MPI-250-12 12 VCT @ 20.8 A 6 V @ 41.7 A
MPI-300-12 12 VCT @ 25.0 A 6 V @ 50.0 A
I'm not trying to shoot you down here, just trying to understand what you mean.
I've read this entire thread and I'd like to thank you for all the helpful information. I'm building a variation of this amp as my first OTL as I'm convinced the topology is the best I've seen.
However, I beg to differ with you assertion that NFB does not reduce output impedance.
You state:
"The power output of this amplifier is unchanged into 4 ohms whether it has zero feedback or 20db. What is different is how the amp tries to respond to the load; it will limit itself to the 40 watts it can make into 4 ohms. Now if NFB **did** lower output impedance, then the 4 ohm power figure would go up".
I think you're mixing up the effects of NFB on maximum power output obtainable with the effects of NFB on output impedance, and they are two different things.
The output impedance won't determine the maximum power output obtainable nor will NFB increase the maximum power output obtainable.
The output impedance is a measure of the volt drop that occurs when the load impedance is changed while the amp is operating at less than rated power.
When the amp is operated well within its power limits, with no NFB and a high output impedance, the frequency response of a system using speakers with a non-constant impedance (IOW, an impedance that changes with frequency) will not be flat. The voltage (and power) for frequencies at which the impedance of the speakers is low will be reduced relative to the frequencies at which the speaker has a higher impedance.
NFB can greatly reduce the variation in frequency response due to the uneven impedance of the speakers by reducing the variation in voltage and power response across the audio spectrum when the amp is operated within the limits of its power output.
If I'm wrong set me straight. This is the way I understand it and have understood it for years.
How many 6H13C grids can be driven with the CF of a 6SN7? Would paralleling the 6SN7 tubes used in the CF ever be necessary to drive 12 grids?
Last edited:
Are the two triodes in a 6AS7 or a 6H13C generally matched or are they just as different in their charactersitics as two triodes in two different tubes?
The 6AS7 was originally intended for use as a series pass element for active voltage regulators. They didn't pay too much attention to balance between sections for that reason: they figured that most users would be paralleling them anyway, with the appropriate current sharing resistors. Also, balance isn't important for voltage regulation use.
As for modern types, I'm not sure how they're made. Once these became known as audio finals, maybe that changed? Ralph definitely favours new production 6AS7s over NOS. Maybe that's one reason? Maybe he'll explicate here for us?
Thanks. I have NOS 6H13C so I'm stuck with those. I puchased 60 tubes and am building a test jig for measuring the bias currents and sorting the tubes.
I think we are on the same page. My point was simply that adding feedback does not increase the output power into lower impedances. So given that, in any other branch of electronics, the output impedance must be unchanged. But in audio, somehow that is not the case. I suspect that it is a matter of convention more than anything else.
You can easily drive 12 power tubes with only one 6SN7 as a cathode follower.
Great. One other question. The VB408 has been discontinued. Is there a substitute you or anyone could recommend? I like the idea of regulation even if it's not absolutely needed.
I have the amp modeled in PSPICE and it's working. I know none of the component values can be read but here's what it looks like with 12 output tubes. I'm getting a clean 60 or 70 watts RMS but the zeners are wrong so the bias is wrong. I need to find a model of a 47 volt zener.
An externally hosted image should be here but it was not working when we last tested it.
When use only unbalanced (RCA) input for Amp drive(same as done in your circuit simulation) , pin 3 of XLR (balanced input) have to be grounded ! In praxis this balanced input `short` is done with short wire jumper placed between pins 1&3 of XLR balanced input.
Best Regards
Best Regards
Beware high voltage zeners drift quite a lot. Perhaps not enough for any real world problems, but expect a volt or two drift.
The batch I have, a box of 100 Sovteks from the same production, there is a difference in them. They are not what I call matched. Tho if you have a good number of them it wont be difficult finding relatively matched sets.
I have found no problems running poorly matched tubes b/c the cathode resistors more or less ensures good current sharing. I wouldn't run them without cathode resistors tho.
Power Limit
I'm running into the same problem with this amp as other simulations I've run. Perhaps it's something I'm doing wrong in modeling the circuit, pehaps it's the model of the 6H13C I'm using, or perhaps it's a limitation of choosing the 6H13C.
I'm finding that the output voltage is limited by the supply voltage, not by the output swing of the driver.
The CF is swinging about 200 Vp-p, but the output tubes clip at around 80 volts p-p into 8 ohms with 160 volt supplies and biased around 10 watts (13 watts is the max plate dissipation). It doesn't matter how many additional tubes are paralleled. Additional tubes do only one thing - lower the output impedance. They do not increase the output voltage or power into a fixed load when you get up to 8 or so tubes. I have 15 tubes for one channel in the latest model.
Also, I'm finding it impossible to operate this amp in class A beyond about one half power unless the tubes are biased such that they would burn up. If they are biased at their maximum plate dissipation they still cut off at full power and the amp operates in AB. I see no way to stop the amp from doing this other than operating it at partial power. The
The 80 volt p-p limitation may be real. That's 80 volts of output swing on tubes powered by a 160 volt source, or 160 volts p-p from a 320 volt source. Can the circlotron output stage voltage exceed 1/2 the supply voltage, or is this some kind of inherent limitation of this topology?
If I raise the supply voltages on the output I can get more output voltage and power with no other modifications. So the driver is doing its job.
Any suggestions or ideas on what I might be doing wrong?
In any event, I can get 200 watts peak (how Atma-Sphere rates their amps), or 100 watts RMS into 8 ohms with a very low output impedance and low distortion.
The MA-2 Mk 3.1 is rated 220 watts/channel into 4, 8 or 16 ohm load before clipping. Notice the lack of the terms "RMS" or "average". The peak power of my simulation is 200 watts, about the same as the MK-2.
The amp also clips gracefully even with 500K feedback resistors.
I'm running into the same problem with this amp as other simulations I've run. Perhaps it's something I'm doing wrong in modeling the circuit, pehaps it's the model of the 6H13C I'm using, or perhaps it's a limitation of choosing the 6H13C.
I'm finding that the output voltage is limited by the supply voltage, not by the output swing of the driver.
The CF is swinging about 200 Vp-p, but the output tubes clip at around 80 volts p-p into 8 ohms with 160 volt supplies and biased around 10 watts (13 watts is the max plate dissipation). It doesn't matter how many additional tubes are paralleled. Additional tubes do only one thing - lower the output impedance. They do not increase the output voltage or power into a fixed load when you get up to 8 or so tubes. I have 15 tubes for one channel in the latest model.
Also, I'm finding it impossible to operate this amp in class A beyond about one half power unless the tubes are biased such that they would burn up. If they are biased at their maximum plate dissipation they still cut off at full power and the amp operates in AB. I see no way to stop the amp from doing this other than operating it at partial power. The
The 80 volt p-p limitation may be real. That's 80 volts of output swing on tubes powered by a 160 volt source, or 160 volts p-p from a 320 volt source. Can the circlotron output stage voltage exceed 1/2 the supply voltage, or is this some kind of inherent limitation of this topology?
If I raise the supply voltages on the output I can get more output voltage and power with no other modifications. So the driver is doing its job.
Any suggestions or ideas on what I might be doing wrong?
In any event, I can get 200 watts peak (how Atma-Sphere rates their amps), or 100 watts RMS into 8 ohms with a very low output impedance and low distortion.
The MA-2 Mk 3.1 is rated 220 watts/channel into 4, 8 or 16 ohm load before clipping. Notice the lack of the terms "RMS" or "average". The peak power of my simulation is 200 watts, about the same as the MK-2.
The amp also clips gracefully even with 500K feedback resistors.
I appreciate the suggestion but in this case the amp is designed so that the XLR pin 3 is not grounded. If pin 3 were grounded with a jumper it would destroy the balanced signal feeding the amp turning it into an unbalanced signal. The design grounds the grid that XLR pin 3 would feed through a 100K resistor when an RCA input is used, and uses the fully balanced signal when an XLR balanced signal is available.
For kicks I grounded pin 3 as you suggested and all it did was degrade the performance of the amp.
For real Amp operation over RCA input pin 3 of XLR input have to be grounded to avoid some slight buzz & hum .
For best Amps sonic performance XLR(balanced) input is the best choice for Amps drive since the whole Amp concept is symmetric and full balanced differential designed from input to output.
Best Regards
My suggestion is not to trust the simulation 100%. As the plate voltage decreases due to high output level, the tube naturally emits relatively less electrons. You will reach a point where adding tubes doese littel to increase output b/c the plate voltage is becoming too low. For more power increase supply voltage, but that means going with colder bias and more classs-B than clas AB. What does the simulation say about grid current at full drive? Perhaps the simulation thinks the tube's cathode is at it's limit as to supplying more electrons?
I doubt the simulation shows the true nature on how the amp clips...The fact that it shows graceful clipping is an indication it says the tubes cannot supply any more current, and has reached it's limit rather gradually.
Increasing the plate voltage increases the power substantially. Unfortunately the output tubes I have can't use a very high B+.
If the tubes could handle a higher B+ the simulation shows that power output above 200 RMS could be achieved. Thinking about this it makes some sense. The way the power supplies are configured the load forms a voltage divider with the internal resistances of the tubes that limits the voltage across the load. It really doesn't matter how much "reserve current" the tubes have the capability to output if there is not enough potential difference to drive the current through the load.
Think about it this way. In the extreme case, the maximum voltage the load will ever see is created when one set of tubes is cutoff and the other set is fully conducting. This essentially becomes a resistive divider consisting of the load in series with the parallel tubes across one of the power supplies.
The internal resistance of the 6H13C is about 250 ohms. With 10 tubes in parallel that reduces to 25 ohms. With an 8 ohm load and a 150 volt power supply, the maximum voltage the load will see is [8/(25+8)] * 150V = 36 volts. Since the voltage can fully reverse polarity, the peak to peak voltage seen by the load is twice this or 72 volts p-p. This is almost identical to the results I'm obtaining. Adding more tubes does increase the power by reducing the parallel resistance of the tubes, but each tube added lowers the impedance less and less than the one before it. 20 tubes would only lower the resistance to 12.5 ohms which would still severely limit the voltage that can reach the load from the supply.
The conclusion I draw is that for high power circlotrons a higher B+ is needed, even for high impedance loads. The reason that more power is not obtained when tubes are added is a simple matter of insufficient supply voltage, not the inability of the tubes to handle the current. If the power supplies were 300 volts and enough tubes were used the power would be limited more by the resistance of the tubes than the power supply voltage.
The ideal OTL tube would have a low plate impedance and could handle high plate voltages so that higher supply voltages could be used.