Hello all,
I'm hoping that someone can through some light on an issue I have following a change in O/T's in my first simple SE amplifier.
I've replaced the old radio O/T's with a pair from VVT in the UK owing to the cores becoming saturated at higher volumes and the impedance mismatch.
However rather than improving it I have lost all of the bass which I find rather confusing.
The new ones have a 5.2K primary which isn't a perfect match for a 6P25 output valve according to the data sheet but it's almost a third of the old O/T's that I had been using which gave a perfectly acceptable sound(to my ears at least) for general listening at moderate levels. The primary inductance is quoted as being 10H.
As VVT have now ceased trading any advice given would be greatly appreciated solving the issue.
I have attached a schematic in case this helps.
Many thanks
Mike
I'm hoping that someone can through some light on an issue I have following a change in O/T's in my first simple SE amplifier.
I've replaced the old radio O/T's with a pair from VVT in the UK owing to the cores becoming saturated at higher volumes and the impedance mismatch.
However rather than improving it I have lost all of the bass which I find rather confusing.
The new ones have a 5.2K primary which isn't a perfect match for a 6P25 output valve according to the data sheet but it's almost a third of the old O/T's that I had been using which gave a perfectly acceptable sound(to my ears at least) for general listening at moderate levels. The primary inductance is quoted as being 10H.
As VVT have now ceased trading any advice given would be greatly appreciated solving the issue.
I have attached a schematic in case this helps.
Many thanks
Mike
There's almost no NFB around that pentode? If you lift one end of the 3.3k, does the gain change much?
The next question will be NFB phase. (Ah, as Steve says.)
The next question will be NFB phase. (Ah, as Steve says.)
Just found *this* in their page:
Page Updated 31 Jul 2017.
FWIW Blue - Red - Brown is classic "American" primary codes, as used in Fender amplifiers and a myriad other Guitar amps.
Brits used others.even in same brand amps, acording to suppliers (Dagnall or Drake for example).
Page Updated 31 Jul 2017.
Maybe you got old colour code ones and miswired transformers?
FWIW Blue - Red - Brown is classic "American" primary codes, as used in Fender amplifiers and a myriad other Guitar amps.
Brits used others.even in same brand amps, acording to suppliers (Dagnall or Drake for example).
Last edited:
Hi, thanks for your replies.
O/T's are drop throughs with solder tags and I've double checked the connections.
The cut off frequency I assume is the reason why the ac signal voltage has dropped by a third at approximately 75Hz compare too 100Hz when measure at the output tube anode.
As for the feedback that was a bit of guess work to be honest as I still can't head around a lot of the maths involved. What would a more appropriate value be?
I can say that I've tried no feedback which hasn't made any noticeable difference. As for the phase I know what it means but how to determine whether it's correct or not is a different matter.
The O/t's were to suit an EL84 which was the closest I could find to suit a Mazda 6P25 or equivalents. Could this have anything to do with it or are they just their design?
To get better results from a 65 year old radio transformer id frustrating to say the least.
Thanks
O/T's are drop throughs with solder tags and I've double checked the connections.
The cut off frequency I assume is the reason why the ac signal voltage has dropped by a third at approximately 75Hz compare too 100Hz when measure at the output tube anode.
As for the feedback that was a bit of guess work to be honest as I still can't head around a lot of the maths involved. What would a more appropriate value be?
I can say that I've tried no feedback which hasn't made any noticeable difference. As for the phase I know what it means but how to determine whether it's correct or not is a different matter.
The O/t's were to suit an EL84 which was the closest I could find to suit a Mazda 6P25 or equivalents. Could this have anything to do with it or are they just their design?
To get better results from a 65 year old radio transformer id frustrating to say the least.
Thanks
VVT look like they are in the process of an orderly shut down, so appear to be honouring their customers with support. Might be worth running it past them?
Could it be the gapping of the transformer for single ended use - it might be designed for a higher current rating? Not sure how much of an effect that could make?
Could it be the gapping of the transformer for single ended use - it might be designed for a higher current rating? Not sure how much of an effect that could make?
I had emailed VVT and this afternoon received a positive reply which is encouraging.
They will look into the issue and if found to be at fault have offered a refund or replacement with an increased iron stack. I assume that the extra iron will increase the inductance thus lowering the cut off frequency.
As for current I have a 40ma anode with 7.5ma screen which isn't far off an EL84.
They will look into the issue and if found to be at fault have offered a refund or replacement with an increased iron stack. I assume that the extra iron will increase the inductance thus lowering the cut off frequency.
As for current I have a 40ma anode with 7.5ma screen which isn't far off an EL84.
Increase feedback (reduce series resistor) until the gain goes down "significantly". To 3/4 or half is OK for guitar.
If the gain goes UP, then squeals full voice, reverse the connections somewhere. (OT primary usually simplest.)
VVT did make very good OPT's, I've used them in several builds with no issues.They have also in the past been brilliant with customer care, but I wouldn't think it's your OPT with a fault, though it can't be ruled out.
To set NFB use a 10 turn pot or preset something like a 10k though any value will do. Set pot to max resistance. If you have a scope connect the OP of your amp to an 8r resistor/dummy load and to the scope and also a DMM/wattmeter or similar to keep an eye on AC RMS. If you have a THD meter this helps too, if not you could use Soundcard Scope,some free PC software that has FFT & THD.
Next with a 1khz sinewave IP of around 500mV to 1v increase volume to about 3v RMS OP (about 1w), start reducing R keeping an eye on scope, OP & THD. Play around till you get low THD but with an acceptable OP power level. You also want to make sure that the amp is stable, any sign of instability or oscillation back off the NFB. You can do this by ear too with a speaker and some music you know well.
Next to find the NFB capacitor. This can involve a lot more work, but a simple way is to feed your amp with a 1khz squarewave with either a capacitance decade box of variable cap across the resistor (or try various fixed caps around the 300 - 1000p range), adjust till you have a good "square" squarewave,no rounding or overshoot. Repeat with a 10khz sqwave. You might find you need to reduce C at 10khz or vice versa to get a good sqwave at both 1kh and 10khz.
Before all this of coarse you can test your amp with no NFB with a sqwave to test your OPT, a sinewave at 20 - 50hz will reveal much too. With a 50hz sine IP increase volume, does the sinewave distort? Same goes for sqwave testing. To make sure it's not another stage in your amp that's reducing the bass response move your scope probe to the grid leak R of the OP stage, repeat test.
Some useful info re squarewave testing here - https://www.kennethkuhn.com › students › ee351 › square_wave_testing.pdf
Andy.
To set NFB use a 10 turn pot or preset something like a 10k though any value will do. Set pot to max resistance. If you have a scope connect the OP of your amp to an 8r resistor/dummy load and to the scope and also a DMM/wattmeter or similar to keep an eye on AC RMS. If you have a THD meter this helps too, if not you could use Soundcard Scope,some free PC software that has FFT & THD.
Next with a 1khz sinewave IP of around 500mV to 1v increase volume to about 3v RMS OP (about 1w), start reducing R keeping an eye on scope, OP & THD. Play around till you get low THD but with an acceptable OP power level. You also want to make sure that the amp is stable, any sign of instability or oscillation back off the NFB. You can do this by ear too with a speaker and some music you know well.
Next to find the NFB capacitor. This can involve a lot more work, but a simple way is to feed your amp with a 1khz squarewave with either a capacitance decade box of variable cap across the resistor (or try various fixed caps around the 300 - 1000p range), adjust till you have a good "square" squarewave,no rounding or overshoot. Repeat with a 10khz sqwave. You might find you need to reduce C at 10khz or vice versa to get a good sqwave at both 1kh and 10khz.
Before all this of coarse you can test your amp with no NFB with a sqwave to test your OPT, a sinewave at 20 - 50hz will reveal much too. With a 50hz sine IP increase volume, does the sinewave distort? Same goes for sqwave testing. To make sure it's not another stage in your amp that's reducing the bass response move your scope probe to the grid leak R of the OP stage, repeat test.
Some useful info re squarewave testing here - https://www.kennethkuhn.com › students › ee351 › square_wave_testing.pdf
Andy.
Last edited:
Thank you for the comprehensive instructions and advice.
I have a scope be it an old CRT one and bench DMM. My frequency generator is some what basic but reasonably accurate I believe.
When I connect my scope to to the output valve grid I get a clean sine wave across a wide frequency range with no voltage drop. Where as I can have a voltage on the anode of 120v down to 100hz after which it has dropped to 80v at 75hz and continues to fall the lower the frequency.
This I take to mean either be a fault with the 6P25's or the O/T's.
I have a scope be it an old CRT one and bench DMM. My frequency generator is some what basic but reasonably accurate I believe.
When I connect my scope to to the output valve grid I get a clean sine wave across a wide frequency range with no voltage drop. Where as I can have a voltage on the anode of 120v down to 100hz after which it has dropped to 80v at 75hz and continues to fall the lower the frequency.
This I take to mean either be a fault with the 6P25's or the O/T's.
You should measure the output ( secondary of the OT ) with a power resistor in place of the speaker , at the reasonable high power near the maximum , this is the correct way , what you did is not very clear ...
No bass with a transformer means low inductance ... maybe much lower than 10H , which is not great anyway .
And compare the measurements with the old transformer .
No bass with a transformer means low inductance ... maybe much lower than 10H , which is not great anyway .
And compare the measurements with the old transformer .
Last edited:
Your welcome.
So,do the same test with NFB and at the same time monitor HT voltage and AC ripple (flick DMM onto AC volts). All evidence so far suggests the VVT OPT is saturating at LF though. Scope the OP of your amp, what happens to the sinewave? Does it start to deform/look like someone's taken a bite out of it?
Couldn't see where the OP says he's testing with a speaker,but your right all tests should be done using a resistor. If you have no power resistor use the biggest 10r resistor you have. You can use a 1/2w 10r in a cup of water, but make sure obviously it's out of the way, not near any high voltage & won't tip over. DO NOT USE WATER WITH DC.
Andy.
That's interesting. Is this with NFB? If so repeat the same test without NFB. If it was without NFB that suggests an impedance mismatch between OP valves and OPT or maybe a power supply issue.
So,do the same test with NFB and at the same time monitor HT voltage and AC ripple (flick DMM onto AC volts). All evidence so far suggests the VVT OPT is saturating at LF though. Scope the OP of your amp, what happens to the sinewave? Does it start to deform/look like someone's taken a bite out of it?
Couldn't see where the OP says he's testing with a speaker,but your right all tests should be done using a resistor. If you have no power resistor use the biggest 10r resistor you have. You can use a 1/2w 10r in a cup of water, but make sure obviously it's out of the way, not near any high voltage & won't tip over. DO NOT USE WATER WITH DC.
Andy.
Last edited:
Diabolical Artificer,
As you know . . .
Quite often a Pentode (and a Beam Power tube too), have much higher plate impedance, rp, than the output transformer primary, RL.
Example, a Pentode with rp = 20k, and a primary with 5k and 10 Henry.
The inductive reactance of 10 Henry is 2500 Ohms at 40Hz.
So, a 5k to 8 Ohm output transformer, with the secondary loaded by an 8 Ohm resistor,
will have a low impedance, RL at 40Hz.
(The 5k primary is in parallel with 2500 Ohms inductive reactance at 40Hz).
That means the amplitude of the bass frequencies will be rolled off, unless you use some kind of negative feedback that involves the signal that is across a winding of the transformer . . . Schade, Global, or Cathode.
If the amplifier does not have Schade, Global, or Cathode negative feedback from an output transformer winding, then . . .
You can Triode Wire the Pentode or Triode Wire the Beam Power tube, to get its plate impedance, rp to be lower (1.5k for example).
1.5k rp will drive the 5k and 10 Henry primary a lot better at 40Hz, than no negative feedback and a pentode/beam power plate impedance, rp of 20k (just an example).
As you know . . .
Quite often a Pentode (and a Beam Power tube too), have much higher plate impedance, rp, than the output transformer primary, RL.
Example, a Pentode with rp = 20k, and a primary with 5k and 10 Henry.
The inductive reactance of 10 Henry is 2500 Ohms at 40Hz.
So, a 5k to 8 Ohm output transformer, with the secondary loaded by an 8 Ohm resistor,
will have a low impedance, RL at 40Hz.
(The 5k primary is in parallel with 2500 Ohms inductive reactance at 40Hz).
That means the amplitude of the bass frequencies will be rolled off, unless you use some kind of negative feedback that involves the signal that is across a winding of the transformer . . . Schade, Global, or Cathode.
If the amplifier does not have Schade, Global, or Cathode negative feedback from an output transformer winding, then . . .
You can Triode Wire the Pentode or Triode Wire the Beam Power tube, to get its plate impedance, rp to be lower (1.5k for example).
1.5k rp will drive the 5k and 10 Henry primary a lot better at 40Hz, than no negative feedback and a pentode/beam power plate impedance, rp of 20k (just an example).
Last edited:
Hadn't done the maths, what you say is probably true, maths not my being strong point, but what is puzzling is why does a reasonable quality OPT sound worse than an OPT pulled from an old radio?
So Z is 5k || 2k5 = 1666, therefore at 40hz we have a 20k 1k6 potential divider, call it 20k/2k for simplicity means we have a Va/Vp voltage of 10v ish @ 40hz; is that right?
It's a reasonable assumption that the radio OPT was cheaply made, had less copper/windings and was probably designed for speech with an abysmal frequency range & having less primary L than yon VVT tfmr; we'd have to measure to be sure.
Andy.
So Z is 5k || 2k5 = 1666, therefore at 40hz we have a 20k 1k6 potential divider, call it 20k/2k for simplicity means we have a Va/Vp voltage of 10v ish @ 40hz; is that right?
It's a reasonable assumption that the radio OPT was cheaply made, had less copper/windings and was probably designed for speech with an abysmal frequency range & having less primary L than yon VVT tfmr; we'd have to measure to be sure.
Andy.
Maybe he did some mistakes that's why I suggested to measure the frequency response of both transformers ...
Of course the radio transformer should be much worse but we don't know for sure what "reasonable quality OPT" he got .
Of course the radio transformer should be much worse but we don't know for sure what "reasonable quality OPT" he got .
Speakers were used only momentarily due to impatience, rest of the time a dummy load
The sound quality is better from the VVT they are just missing the bottom end which I expected would be improved not reduced.
As soon as I have done a comparison I will report back.
However I can advise that VVT having checked their records and my schematic are making FOC replacements for me.
The sound quality is better from the VVT they are just missing the bottom end which I expected would be improved not reduced.
As soon as I have done a comparison I will report back.
However I can advise that VVT having checked their records and my schematic are making FOC replacements for me.
Now that is impressive customer service - a company in the process of closing down offering free replacement transformers. I wish that VVT could stay in business!
Finally got the opportunity to swap over one of the OT's for a comparison which
has left me even more confused.
The old OT has a winding ratio of 40:1 which gives a primary impedance of
40*40*8 = 12800 ohms which is not surprising as it was no doubt wound for a
much lower impedance speaker than an 8 ohm one.
As far as I was aware and correct me if I am wrong a higher load impedance reduces
the output.
The output voltage at the speaker terminals is higher for for the old OT.
In fact it maxes out a 9v which I make to be 10.1W which is basically the
quoted max dissipation for a 6P25 where as for the VVT it's 4.2w. this fits in
with the quoted typical operating conditions.
This is without any feedback and yes, plenty of distortion.
At lower frequencies and realistic listening levels the output drops drastically
to the point of being non existent.
This is the reason why the old OT gives a better bass response but how it achieves it
makes no sense to me. Are there circumstances where a higher load impedance increase power output?
Thanks
has left me even more confused.
The old OT has a winding ratio of 40:1 which gives a primary impedance of
40*40*8 = 12800 ohms which is not surprising as it was no doubt wound for a
much lower impedance speaker than an 8 ohm one.
As far as I was aware and correct me if I am wrong a higher load impedance reduces
the output.
The output voltage at the speaker terminals is higher for for the old OT.
In fact it maxes out a 9v which I make to be 10.1W which is basically the
quoted max dissipation for a 6P25 where as for the VVT it's 4.2w. this fits in
with the quoted typical operating conditions.
This is without any feedback and yes, plenty of distortion.
At lower frequencies and realistic listening levels the output drops drastically
to the point of being non existent.
This is the reason why the old OT gives a better bass response but how it achieves it
makes no sense to me. Are there circumstances where a higher load impedance increase power output?
Thanks
Output power from a single-ended amplifier is max 50% plate dissipation , so only 5W is realistic from 1 tube dissipating 10W .
Maximum output power is allways at optimum Ra , in this case around 5K , not higher .
So consider this when testing ... if the results don't follow this rules there is a mistake .
Maximum output power is allways at optimum Ra , in this case around 5K , not higher .
So consider this when testing ... if the results don't follow this rules there is a mistake .
Last edited: