Could ask either forum...
Peavey VB2 bass amp with 6 EL34 (mono).
I replaced the output transformer with one for a Trace Elliot V6 amp that used 6 KT88 and I replaced all the EL34 output tubes. The feedback wire is not currently connected to the output transformer, as I was unsure of the correct polarity with the different new transformer.
Hasn't run in many years.
I haven't done much of this kind of work in a long long time. So I'll need advice along the way. First I guess is my test equipment.
1) I don't have a 'bench-type' signal generator, but my Behringer digital EQ/RTA has a sine/white/pink generator with a balanced output, and nice discrete steps for output level and frequency. I imagine I should make some attempt to protect it's output if I'm going to be sticking it into various places in tube amps to insert signal along the path. Should I put the Behringer output into a small power amp so I start with excess voltage and current capability, then drop it with resistors in an "H" pad and finally output that to some coupling caps and some leads? Should I just buy a cheap used audio bench signal generator?
2) I got a very decent used Tektronix analog scope! But it only has X10 probes. Do I really need X100 probes? I don't want to hurt it!
3) I don't have a dummy load instead of a speaker; I used to have a Scholz power soak but they don't really handle much power. Any good recommendations? I'd like to make something with some jacks for various load impedances which I could use both as a dummy load and as a power soak.
Peavey VB2 bass amp with 6 EL34 (mono).
I replaced the output transformer with one for a Trace Elliot V6 amp that used 6 KT88 and I replaced all the EL34 output tubes. The feedback wire is not currently connected to the output transformer, as I was unsure of the correct polarity with the different new transformer.
Hasn't run in many years.
I haven't done much of this kind of work in a long long time. So I'll need advice along the way. First I guess is my test equipment.
1) I don't have a 'bench-type' signal generator, but my Behringer digital EQ/RTA has a sine/white/pink generator with a balanced output, and nice discrete steps for output level and frequency. I imagine I should make some attempt to protect it's output if I'm going to be sticking it into various places in tube amps to insert signal along the path. Should I put the Behringer output into a small power amp so I start with excess voltage and current capability, then drop it with resistors in an "H" pad and finally output that to some coupling caps and some leads? Should I just buy a cheap used audio bench signal generator?
2) I got a very decent used Tektronix analog scope! But it only has X10 probes. Do I really need X100 probes? I don't want to hurt it!
3) I don't have a dummy load instead of a speaker; I used to have a Scholz power soak but they don't really handle much power. Any good recommendations? I'd like to make something with some jacks for various load impedances which I could use both as a dummy load and as a power soak.
Last edited:
As far as I know, the easy way to check the polarity of the feedback connection at the transformer is to try it and see. (There are some very knowledgeable people on here who might suggest a better way.) Do this after you have your dummy load set up, as it would be very loud through a speaker if you happen to connect to the wrong polarity and it oscillates!
I would strongly recommend wiring up a 'lamp limiter' for initial testing. There are threads on here, and elsewhere, which can explain it.
Also before going to signal generator and 'scope', measure the DC voltages at the important points in the amp, to see if they are reasonable. Check the B+ (or HT) at each stage, check the voltage at each anode and cathode and also at each grid. This is very dangerous, so check the safety instructions here, and elsewhere. For me the most important rule is to only use one hand (keep the other hand in your pocket).
There may be no need to inject test signals into the amp, or use the scope. If all the DC levels look reasonable, just try it with a guitar.
If you do want to use test signals, I would imagine the Behringer puts out a line level signal, which would just about be OK to inject into the amp for test purposes (it may overdrive sensitive input stages and would provide a weak drive to output stages. The important thing is to put capacitors in series in the test lead (0.1uF on both wires). This is to protect the Behringer from possible DC levels in the amp, and to protect the amp against the possibility that one of the wires is grounded somehow (I know it shouldn't be, as its supposed to be balanced).
I think the X10 probes will be OK for most points in the amp. (The anodes of the output valves will be the place where the AC voltage is too high for them.)
You can buy 8 ohm 100 watt resistors, which in series/parallel give you a 400 watt load. They can still get pretty hot though, so bolting to a metal chassis or proper heat sink is recommended. I guess your amp will output about 150 watts? You could also arrange two of the 8 ohms in series to give a 16 ohm 200 watt load. I wouldn't run them too close to rating though, unless the heat sinking is really good.
Hope this helps, and if I've forgotten something I'm sure other members will chip in.
I would strongly recommend wiring up a 'lamp limiter' for initial testing. There are threads on here, and elsewhere, which can explain it.
Also before going to signal generator and 'scope', measure the DC voltages at the important points in the amp, to see if they are reasonable. Check the B+ (or HT) at each stage, check the voltage at each anode and cathode and also at each grid. This is very dangerous, so check the safety instructions here, and elsewhere. For me the most important rule is to only use one hand (keep the other hand in your pocket).
There may be no need to inject test signals into the amp, or use the scope. If all the DC levels look reasonable, just try it with a guitar.
If you do want to use test signals, I would imagine the Behringer puts out a line level signal, which would just about be OK to inject into the amp for test purposes (it may overdrive sensitive input stages and would provide a weak drive to output stages. The important thing is to put capacitors in series in the test lead (0.1uF on both wires). This is to protect the Behringer from possible DC levels in the amp, and to protect the amp against the possibility that one of the wires is grounded somehow (I know it shouldn't be, as its supposed to be balanced).
I think the X10 probes will be OK for most points in the amp. (The anodes of the output valves will be the place where the AC voltage is too high for them.)
You can buy 8 ohm 100 watt resistors, which in series/parallel give you a 400 watt load. They can still get pretty hot though, so bolting to a metal chassis or proper heat sink is recommended. I guess your amp will output about 150 watts? You could also arrange two of the 8 ohms in series to give a 16 ohm 200 watt load. I wouldn't run them too close to rating though, unless the heat sinking is really good.
Hope this helps, and if I've forgotten something I'm sure other members will chip in.
There is a way of checking if global NFB is safe to apply; instead of the low values usually found on the o/p secondary to the global NFB loop, often I use a much higher value, 47K-100K and if the polarity is incorrect, the amp gain will slightly increase. Do no more as it is just an indicator !
Richy
Richy
Thanks, especially for the clue about the neg feedback; I was expecting to try it powered with lower voltage from the VARIAC, like about 90 volts, so that when it would oscillate it would be less likely to have enough power to damage anything.
Should I assume that a typical KT88's design drove more current into a lower-impedance load than a typical EL34 design? Or does a typical KT88 design use more B+ voltage? I assumed the transformer for the KT88's might have a lower turns ratio, so I attached the transformer lead labeled '8 ohms' to the '16 ohms' speaker terminal at the back of the amp, and the 4-ohm tranny lead to the '8 ohms' speaker terminal at the back of the amp, etc. I'm not sure whether this was right.
I figured that once it was running I'd test into various dummy loads and just see what load gave the most output in calculated watts before visible waveform distortion. Is this a reasonable approach, as long as I make sure I'm looking for output distortion and not seeing effects of the preamp?
Should I assume that a typical KT88's design drove more current into a lower-impedance load than a typical EL34 design? Or does a typical KT88 design use more B+ voltage? I assumed the transformer for the KT88's might have a lower turns ratio, so I attached the transformer lead labeled '8 ohms' to the '16 ohms' speaker terminal at the back of the amp, and the 4-ohm tranny lead to the '8 ohms' speaker terminal at the back of the amp, etc. I'm not sure whether this was right.
I figured that once it was running I'd test into various dummy loads and just see what load gave the most output in calculated watts before visible waveform distortion. Is this a reasonable approach, as long as I make sure I'm looking for output distortion and not seeing effects of the preamp?
It seems to be difficult to find any parameters for the Peavey VB2 and Trace Elliot V6 output transformers, but I looked up the anode resistance of the EL34 (ra = 15K) and the KT88 (ra = 12K). Typically, the 'optimal' design load impedance is in proportion to anode resistance, which would suggest that the primary impedance of these two OTs could be reasonably close. (This is just a guestimate, however.)
I think your approach of trying different loads, looking for maximum output seems like a good idea.
I think your approach of trying different loads, looking for maximum output seems like a good idea.
Last edited:
That helps; I could also look on the schematics for the test voltage value at the anodes on each. I'll say one thing, the Trace Elliot V6 output transformer for 6 KT88's is big, in fact it's the same size core (but different windings) as the V8 for 8 KT88's. I expect it to be forgiving and sound pretty good. The feedback wire on the VB-2 is on one of the transformer secondary taps (I'd have to look at which impedance tap) but if that's wrong I have to switch the primaries if I want it scaled the same. Does that mean it's an "Ultralinear" design? I think I'll listen to it a few ways and maybe even put it on a switch. I know a lot of people use a feedback switch on small guitar amps, but I don't have a clue what a bigger more powerful bass amp sounds like without a negative feedback loop connected.
For ultra-linear, there will be extra taps on the primary side which are connected to the output tube screen grids, possibly via screen stopper resistors. If it is normal pentode mode, the screen grids connect (again possibly via screen stoppers) to a B+ supply.
A feedback switch, or even a pot. if you have room for it, could be good. I don't have any experience of no-NFB bass amps, but I know what it does in a guitar amp. Without NFB, the output of the amp (normal pentode mode) tends to follow the impedance curve of the speaker, so you get increased output near the speaker resonant frequency and increased highs (where the speaker impedance also rises). Output stage distortion is also more gradual in its onset.
Without NFB, you smooth out the speaker resonance and remove the added top end brightness. Power stage distortion also has a more abrupt onset.
Its worth trying to see which you prefer, but I think for a powerful bass amp the uncontrolled low frequency speaker resonance (without NFB) may give too much 'boom'. Of course it would depend on where the speaker resonant frequency is (or frequencies, if you mix 18" and 10" for example).
A feedback switch, or even a pot. if you have room for it, could be good. I don't have any experience of no-NFB bass amps, but I know what it does in a guitar amp. Without NFB, the output of the amp (normal pentode mode) tends to follow the impedance curve of the speaker, so you get increased output near the speaker resonant frequency and increased highs (where the speaker impedance also rises). Output stage distortion is also more gradual in its onset.
Without NFB, you smooth out the speaker resonance and remove the added top end brightness. Power stage distortion also has a more abrupt onset.
Its worth trying to see which you prefer, but I think for a powerful bass amp the uncontrolled low frequency speaker resonance (without NFB) may give too much 'boom'. Of course it would depend on where the speaker resonant frequency is (or frequencies, if you mix 18" and 10" for example).
My scope came with a pair of x10 probes rated at 400V and I blew one up at 450v. The thing was, I was looking at a pulse with a risetime of about 1uS. The voltage rating is only for DC and low frequencies. At HF you have to de-rate them. Also, x100 probes will have a 100Meg input impedance and probably lower capacitance. Maybe use the x100 probes for anything over say 50V signal. If you use them on too small a signal you have to crank the scope gain way up and it may introduce it's own noise onto the display.
This really belongs in the Instrument & Amps forum so I'll move it there. You may get some additional useful input there.
- Status
- Not open for further replies.