I have a pair of Quad IIs I bought needing some work. They were missing a few parts, but restored them back to original spec, with some very minor mods as mentioned by Keith Snook and Patrick Turner:
The amplifier sounds great when used with a source like a bluetooth adaptor, but if I want to listen to CDs I need to use a valve preamp (Opus Consonance) since there is no gain/volume control on the quads.
If the CD is playing, the amp is on, and I turn on the preamp, I briefly hear some music and a bit of motorboating, then it goes quiet - no signal in the amp.
If the amp is on and the preamp is on, and I attempt to break and remake the input connection, there is a terrifically loud mains hum sound. The first time that happened, the Quads went quiet, and I went to bed thinking I had fried the OPTs. However following morning, they were happy again.
I guess the significant thing is that the problem is identical with both amps connected - this points to an issue with the preamp, or an interworking issue. Both Quads have the identical build, apart from some coupling caps in the quads which were 1000V in one instead of 630V.
I also have a Tubelab SSE, and that plays fine with the preamp, although it did sound a bit thin today, compared with linking it directly to the CD player (the SSE has a volume control).
Does anyone have any thoughts I was wondering if the coupling caps on the preamp outputs were the problem - they are Auricaps, 2uF.
- Seperate output tube cathode bias resistors, 390R, bypassed with 150uF
- 0.1uF decoupling capacitor on the input RCA, to protect from DC
- 47uF power supply electrolytics instead of 16uF. I use NOS 5R4GYB rectifier and CL-140 NTC on the primary side to minimise the 'bump' on start up, with no issues.
- 5B/255M output tubes with 100R on the screens, so current can be measured. B+ is 300V, so it is within spec. This tube looks like a KT66 at these voltages.
The amplifier sounds great when used with a source like a bluetooth adaptor, but if I want to listen to CDs I need to use a valve preamp (Opus Consonance) since there is no gain/volume control on the quads.
If the CD is playing, the amp is on, and I turn on the preamp, I briefly hear some music and a bit of motorboating, then it goes quiet - no signal in the amp.
If the amp is on and the preamp is on, and I attempt to break and remake the input connection, there is a terrifically loud mains hum sound. The first time that happened, the Quads went quiet, and I went to bed thinking I had fried the OPTs. However following morning, they were happy again.
I guess the significant thing is that the problem is identical with both amps connected - this points to an issue with the preamp, or an interworking issue. Both Quads have the identical build, apart from some coupling caps in the quads which were 1000V in one instead of 630V.
I also have a Tubelab SSE, and that plays fine with the preamp, although it did sound a bit thin today, compared with linking it directly to the CD player (the SSE has a volume control).
Does anyone have any thoughts I was wondering if the coupling caps on the preamp outputs were the problem - they are Auricaps, 2uF.
Always start the system in CDP, preamp, power amp order. When powering down, the order is reversed.
The order given keeps power on transients from causing havoc.
The order given keeps power on transients from causing havoc.
You should check the output from the pre-amp has no standing voltage on it, just in case.
If it is a 'totem pole' configuration to the 2uF cap it will have a voltage until the cap is charged up?
If it is a 'totem pole' configuration to the 2uF cap it will have a voltage until the cap is charged up?
Thanks for the suggestions. I could do with a schematic for the preamp - I could not find one online. It has a lot of tubes considering it has no phono input or tone control - a rectifier and 4 double triodes.
More and more I wish I had a basic oscilloscope, and the skills to go with it. Is there a good tutorial anyone can recommend?
More and more I wish I had a basic oscilloscope, and the skills to go with it. Is there a good tutorial anyone can recommend?
You can use the soundcard of your PC or laptop to analyze audio. Or you can use an external USB sound card if you need better performance, like for example a lower noise floor. You can measure with a scope, spectrumanalyzer, bode plotter, etc.. Knowing Google is our friend, see for example:
Analyzer Software Overview
Regards, Gerrit
Analyzer Software Overview
Regards, Gerrit
Can I just check that mod 2) 0.1uF decoupling capacitor on the input RCA, to protect from DC. You have a resistor on the grid to ground should be 1.5M could be a lot lower. Just wonder if there's a large DC voltage on the pre-amp output as Alan suggested. Check with a meter.
You may not need a pre-amp, a simple 100k log pot between the CD player and the amp may do.
You may not need a pre-amp, a simple 100k log pot between the CD player and the amp may do.
Last edited:
oops!
Well, one of the Quads died. (It was a 'good age', as my father hates anyone to say).
When I added the decoupling cap, I had overlooked the need for another resistor for the input tube grid bias, so the EF86 grid was hanging of the 0.1uF cap and nothing else. (I'd like to think I've learnt something since I made this improvement, so would not miss something so obvious).
That explains the strange behaviour I have been having, which I think is two problems.
1) Reading through Patrick Turner's site, and the mods he did to a pair of Quads, he described the sort of LF instability that I had when messing with the inputs, and linked that to seperate cathode bias.
2) Finally getting to grips with my new oscilloscope, and I can see a lot of oscillation in the wave from the EF86 anode, with a 1kHz test signal. That is pointing to HF oscillation (around 75kHz).
3) The tube must have suffered having had the grid floating close to cathode potential for so long. At the least, the transient from disconnecting the input must have been able to trigger the LF instability.
The odd thing was that the amps would stop working, then start again a day later, and this must have been due to a negative charge that was persistent on one side of the decoupling cap.
Anyway, fixed the issue. I also added a 22pF/22K across the supply to the output tube grid coupling caps, which Patrick Turner found necessary to tame HF.
With new EF86's it sounds fine again.
I tested the old EF86's, and the one that was connected to the input showed excessive cathode current. Tubelab recently described the mechanism where the grid captures positive ions from the cathode, causing it to conduct more over time.
So I suppose my lovely low use Philips Miniwatt EF86 is off to join that massive amplifier in the sky :-(
Richard
Well, one of the Quads died. (It was a 'good age', as my father hates anyone to say).
When I added the decoupling cap, I had overlooked the need for another resistor for the input tube grid bias, so the EF86 grid was hanging of the 0.1uF cap and nothing else. (I'd like to think I've learnt something since I made this improvement, so would not miss something so obvious).
That explains the strange behaviour I have been having, which I think is two problems.
1) Reading through Patrick Turner's site, and the mods he did to a pair of Quads, he described the sort of LF instability that I had when messing with the inputs, and linked that to seperate cathode bias.
2) Finally getting to grips with my new oscilloscope, and I can see a lot of oscillation in the wave from the EF86 anode, with a 1kHz test signal. That is pointing to HF oscillation (around 75kHz).
3) The tube must have suffered having had the grid floating close to cathode potential for so long. At the least, the transient from disconnecting the input must have been able to trigger the LF instability.
The odd thing was that the amps would stop working, then start again a day later, and this must have been due to a negative charge that was persistent on one side of the decoupling cap.
Anyway, fixed the issue. I also added a 22pF/22K across the supply to the output tube grid coupling caps, which Patrick Turner found necessary to tame HF.
With new EF86's it sounds fine again.
I tested the old EF86's, and the one that was connected to the input showed excessive cathode current. Tubelab recently described the mechanism where the grid captures positive ions from the cathode, causing it to conduct more over time.
So I suppose my lovely low use Philips Miniwatt EF86 is off to join that massive amplifier in the sky :-(
Richard
I have been thinking about that, since i have a few, but I would like to make it reasonably easy to reverse. With the old smoothing cap box no longer fitted, I do have some real estate up by the transformer, so potentially could split the filament supply and have some sort of 'jumper' for the input tubes, with a 3,3R 2W resistor being shorted when an EF86 is fitted.
A resistor in series : from tubular-well.se :
Ett motstånd på 6.4 ohm i serie gör den användbar med 6.3V glöd.
( a resistor of 6.4 ohm in series is whats needed to use PF86 instead of EF86,
1w is enough)