Ccschua,
No problem about R3. Remember (if I understand you correctly) that the NFB resistor is not isolated. One side is connected to R3 (680 ohm) and the other to the transformer voice coil winding. The dc resistance of that will be about 0 ohm. So you are actually measuring not over the NFB resistor, but from the top of R3 to its grounded bottom through the transformer secondary. Thus you are reading R3. You will have to disconnect the feedback resistor(s) to be able to measure them.
No, I fear the filter cap and rectifiers are not going to affect the soundstage etc. They do dc things only. You did not state whether the change of 220K down did have an effect. As said before, that high value resistor chain will lead to inaccuracy e.g. with different tubes. I think one must change the 220K for a lower resistance voltage divider giving a more stable B3 voltage. I will see if I can help before the weekend.
If you look at the frequency response I gave earlier for the front end, there should be no problem with bass etc. That leaves the OPT only, unless you can trace a wiring fault somewhere. A pity you do not have a signal generator to test things. Is it possible to get hold of one?
The change in heater voltage is only 2½%. That is easily the change in mains voltage, probably. It is not important.
No problem about R3. Remember (if I understand you correctly) that the NFB resistor is not isolated. One side is connected to R3 (680 ohm) and the other to the transformer voice coil winding. The dc resistance of that will be about 0 ohm. So you are actually measuring not over the NFB resistor, but from the top of R3 to its grounded bottom through the transformer secondary. Thus you are reading R3. You will have to disconnect the feedback resistor(s) to be able to measure them.
No, I fear the filter cap and rectifiers are not going to affect the soundstage etc. They do dc things only. You did not state whether the change of 220K down did have an effect. As said before, that high value resistor chain will lead to inaccuracy e.g. with different tubes. I think one must change the 220K for a lower resistance voltage divider giving a more stable B3 voltage. I will see if I can help before the weekend.
If you look at the frequency response I gave earlier for the front end, there should be no problem with bass etc. That leaves the OPT only, unless you can trace a wiring fault somewhere. A pity you do not have a signal generator to test things. Is it possible to get hold of one?
The change in heater voltage is only 2½%. That is easily the change in mains voltage, probably. It is not important.
Johan Potgieter said:
Feel so ashame. Reading the resistor codes gives me High NFB = 120K and Low NFB = 150K. I will dismantle and confirm that reading.
I am still awaiting the parts from Partsconnexion. As I believe in 'Mills' resistor for power job. However, I will try over the weekend using a 'Kiwami 180k' for the job. When I say soundstage and bass, I am being bias because i compare that little tiny 6v6 tube amp vs the muscular 'unison research unico 80W amp'
The OPT is only a tiny 4"x4" x 4" height.
I was thinking of getting a DSO with built in signal generator such as PICO. However the costs of shipping is half the price of machine.
On one occasion, the 6v6GT tubes glows like red hot metal, and I thought why I have additional filament light bulb over my place. I measured and the filament has gone up to 6.5V. This is not critical now as I am using a step down transformer to bring it back to 6.3V filament.
BEWARE!
When you say the 6V6GT tubes glow like red hot metal, do you mean the (heated) cathode, or the anodes? If the latter, shut down pronto because the tube may be ruined shortly. The latter means too much anode current; that has nothing to do with heater voltage. In that case you either have a leaky coupling capacitor or lack of bias for some reason.
Meanwhile I have reconstructed the first stage and wonder. I get between 90 - 100V on anode with tubes of varying age, also different manufacturers. But I find tubes of 'JJ' make to give 80% of the specified current. I would suggest careful measurement. Is your meter a digital with 10meg input? Then with 298V at B2 point, one gets 155V at B3, and 93V on the anode. I tried various makes of tube, (found JJ to be 84% only of the EH) and not one was higher than 102V on anode. That is still passable. Thus I do not understand why your Va goes down to below 90V.
I am then not going to suggest a resistive divider for B3 as said before. We must first determine why you are not reaching 90V on the first anode. Again, it is rather difficult to pin-point a fault with only e-mail at one's disposal. You do have a capacitor from B3 to common? You must have some 155V at B3, otherwise that electrolytic may be leaking. Please comment on what exactly you get there, with what kind of meter input resistance before we can go further.
When you say the 6V6GT tubes glow like red hot metal, do you mean the (heated) cathode, or the anodes? If the latter, shut down pronto because the tube may be ruined shortly. The latter means too much anode current; that has nothing to do with heater voltage. In that case you either have a leaky coupling capacitor or lack of bias for some reason.
Meanwhile I have reconstructed the first stage and wonder. I get between 90 - 100V on anode with tubes of varying age, also different manufacturers. But I find tubes of 'JJ' make to give 80% of the specified current. I would suggest careful measurement. Is your meter a digital with 10meg input? Then with 298V at B2 point, one gets 155V at B3, and 93V on the anode. I tried various makes of tube, (found JJ to be 84% only of the EH) and not one was higher than 102V on anode. That is still passable. Thus I do not understand why your Va goes down to below 90V.
I am then not going to suggest a resistive divider for B3 as said before. We must first determine why you are not reaching 90V on the first anode. Again, it is rather difficult to pin-point a fault with only e-mail at one's disposal. You do have a capacitor from B3 to common? You must have some 155V at B3, otherwise that electrolytic may be leaking. Please comment on what exactly you get there, with what kind of meter input resistance before we can go further.
Pls dont laugh. I am just using this Sanwa PM3 meter.
I cant seem to find the input resistance of the meter.
The warm glow has happens at mid part (not sure if it is cathode or anode) anode should be on the outer ring right.
After I switch back to using step down tranny from 240V to 220V works fine. I have no problem. I have decided to open up the pcb and confirm the cap goes to where.
I cant seem to find the input resistance of the meter.
The warm glow has happens at mid part (not sure if it is cathode or anode) anode should be on the outer ring right.
After I switch back to using step down tranny from 240V to 220V works fine. I have no problem. I have decided to open up the pcb and confirm the cap goes to where.
Attachments
Yes, the tube in the middle (about 2 - 3mm diameter) is the cathode. That always heats up to a dull red to emit the electrons used for amplification. The outer 'tube', perhaps oval in shape and some 15mm diameter is the anode. That should not heat to show even a very dull red - easy to see in the dark.
No need to laught at your meter, it can do more than the normal digital meter! The internal resistance is shown under "Impedance", the right-hand column. It is 10meg, thus fine. The move 240-220V should not make such a large difference, but OK if it works better. Please tell what you measure at 'B3".
No need to laught at your meter, it can do more than the normal digital meter! The internal resistance is shown under "Impedance", the right-hand column. It is 10meg, thus fine. The move 240-220V should not make such a large difference, but OK if it works better. Please tell what you measure at 'B3".
As I was curious about why the cathode glow brightly, I switch back R12 to 22k with the input voltage at 240V.
So I did a measurement and this is what I get
V +B1 = 331V
V + B2 313V
V +B3 = 108V
V3a Anode = 260.6V
V3b anode 258V
V3 common cathode 74V
V1 anode 63.5V
V1 cathod 0.47V
V4 V5 common cathode 23.5V
V filament AC = 6.48V (Cant figure out why this changes)
V filament dc = 6.4V
Here is another view of tube amp
So I did a measurement and this is what I get
V +B1 = 331V
V + B2 313V
V +B3 = 108V
V3a Anode = 260.6V
V3b anode 258V
V3 common cathode 74V
V1 anode 63.5V
V1 cathod 0.47V
V4 V5 common cathode 23.5V
V filament AC = 6.48V (Cant figure out why this changes)
V filament dc = 6.4V
Here is another view of tube amp
Getting a little difficult now.
But first you need a compliment on your very clear pictures and schematics!
Now then. I kept asking for B3, and there is a difficulty there. If you calculate the current through that unnamed 220K between B2 and B3 (simple Ohms Law: I = V/R), you get 0,85 - 0,93mA with the two mains feeds. The same current flows through R4 - nothing else attached there - but one gets only 0,42mA there. Where does the other 0,47mA go? I expected an electrolytic capacitor from B3 to common; I don't see one (it is probably too large to be under the board).
What I do notice on the board, just above the second control from the left, is a circle half filled in. That is usually the symbol for an electrolytic. I just cannot make out the annotation, but judging from what I can see (partly obscured by two yellow wires), is that that is probably the place for the missing capacitor - all this if the larger 100K resistor running vertical just above the white wire there is in fact R4. OK, too much guessing; you will have to investigate regarding that capacitor. If it is there somewhere I expect it is leaking. Do go back and find that in the identical circuit I hooked up, B3 was some 157V (I did not recheck now). That would be the expected voltage.
That is why the V1 anode voltage is low and differs from my result of about 90V. You must find what is going on there. With normal design the 220K would be a decoupling resistor feeding to the 'missing' cap, from which junction you then have R4 as the normal tube anode load resistor. If there is no 'missing' capacitor, the 'disappearing' current makes no sense - it is also non-sensical to have mounted R5.C2 over part of the anode load, if that load was meant to be all of 220K + 100K. Thus, I am trying to say, there must be a capacitor from B3 - common (about 22µF would suffice).
Regarding the slight increase in heater voltage I would not worry, although 6,3V would actually result in a somewhat longer tube life.
Also late here now; perhaps you can attend to the above and we will get there yet!
Regards
But first you need a compliment on your very clear pictures and schematics!
Now then. I kept asking for B3, and there is a difficulty there. If you calculate the current through that unnamed 220K between B2 and B3 (simple Ohms Law: I = V/R), you get 0,85 - 0,93mA with the two mains feeds. The same current flows through R4 - nothing else attached there - but one gets only 0,42mA there. Where does the other 0,47mA go? I expected an electrolytic capacitor from B3 to common; I don't see one (it is probably too large to be under the board).
What I do notice on the board, just above the second control from the left, is a circle half filled in. That is usually the symbol for an electrolytic. I just cannot make out the annotation, but judging from what I can see (partly obscured by two yellow wires), is that that is probably the place for the missing capacitor - all this if the larger 100K resistor running vertical just above the white wire there is in fact R4. OK, too much guessing; you will have to investigate regarding that capacitor. If it is there somewhere I expect it is leaking. Do go back and find that in the identical circuit I hooked up, B3 was some 157V (I did not recheck now). That would be the expected voltage.
That is why the V1 anode voltage is low and differs from my result of about 90V. You must find what is going on there. With normal design the 220K would be a decoupling resistor feeding to the 'missing' cap, from which junction you then have R4 as the normal tube anode load resistor. If there is no 'missing' capacitor, the 'disappearing' current makes no sense - it is also non-sensical to have mounted R5.C2 over part of the anode load, if that load was meant to be all of 220K + 100K. Thus, I am trying to say, there must be a capacitor from B3 - common (about 22µF would suffice).
Regarding the slight increase in heater voltage I would not worry, although 6,3V would actually result in a somewhat longer tube life.
Also late here now; perhaps you can attend to the above and we will get there yet!
Regards
Hi Johan,
Thanks for your patience with me. I hope not to miss your pertinent points.
You are right on the voltage between R1. I probe the differential voltage and I got 0.03V, or probably negligible difference.
The electrolytics cap are not accessible, need time to unsolder the wires. there is no point to further guess what the cap is doing unless I dismantle the pcb.
I wonder if the 23V is too high for the 6v6GT common cathode. Should this be reduced to 21V or rather this is all dependent on what we have at V1 and V3 ?
How to check for a leaking capacitor? Can I just probe the terminals without dismantling the capacitor and make sure the value is what is stated or need to unsolder the cap.
rgds
Thanks for your patience with me. I hope not to miss your pertinent points.
You are right on the voltage between R1. I probe the differential voltage and I got 0.03V, or probably negligible difference.
The electrolytics cap are not accessible, need time to unsolder the wires. there is no point to further guess what the cap is doing unless I dismantle the pcb.
I wonder if the 23V is too high for the 6v6GT common cathode. Should this be reduced to 21V or rather this is all dependent on what we have at V1 and V3 ?
How to check for a leaking capacitor? Can I just probe the terminals without dismantling the capacitor and make sure the value is what is stated or need to unsolder the cap.
rgds
I am puzzled by your reference to R1 (470K) and the voltage across it. I cannot find where I spoke of R1 or any voltage there? Our concern is with the above normal voltage drop across the 220K between B2 and B3 (i. e. low B3), and what is happening at B3.
The 'missing' cap, if indeed it is, will be about the size of the other larger electrolytics. I am not sure that that will be mounted under the PC.
The 6V6 cathode voltage of either 21V or 23V is in order; don't worry about those.
The 'missing' cap, if indeed it is, will be about the size of the other larger electrolytics. I am not sure that that will be mounted under the PC.
The 6V6 cathode voltage of either 21V or 23V is in order; don't worry about those.
Hi Johan,
here it is. The 100uF/100V is for the 6v6GT common cathode. Meaning there is no bypass cap at V1 cathode.
at B2+, only 1 cap. The other cap is at the B3+ position as shown in the circuit.
Hope this is clear. I will put in 18k, and perform a measurement. at the same time, I will ensure the soldering is spot on.
No other cap below the PCB. i.e. The empty cap at the yellow wire is not there.
Pls note R12 is still 22k. I will put in 18k at R12 and no other change. Measurement will follow.
Tks for your perseverence for my long windedness.
here it is. The 100uF/100V is for the 6v6GT common cathode. Meaning there is no bypass cap at V1 cathode.
at B2+, only 1 cap. The other cap is at the B3+ position as shown in the circuit.
Hope this is clear. I will put in 18k, and perform a measurement. at the same time, I will ensure the soldering is spot on.
No other cap below the PCB. i.e. The empty cap at the yellow wire is not there.
Pls note R12 is still 22k. I will put in 18k at R12 and no other change. Measurement will follow.
Tks for your perseverence for my long windedness.
Ah! We are getting there!
But first, forget about measuring at R11 by connecting a meter directly there. The resistances there are high, and even a 10meg voltmeter load will alter the voltages enough to be misleading. (Always when measuring, one must keep meter-loading in mind. Even the best normal meters have an input resistance, and it acts like a resistance added where you measure. In the case of an LTP such as here, and when one is sure the LTP is working more or less normal, the combined cathode voltage will reflect the preceeding anode voltage plus a few volt cathode-g1, which one can guess. The tied cathodes are a low impedance point, which a meter is unlikely to disturb. You actually have no voltage drop across R11; it is the meter.)
But then, on to +B3. That C13 which we have finally identified; unsolder it! I realise it is difficult since you must get to the 'underside' of the PC. Also disconnect/reconnect of wires. We need to know the +B3 voltage without C13, as explained previously. I hope it will be in the region of 150V and that the voltage over R12 will now go to 90 - 100V or so. Please note: This is not a conditon in which to try and listen; we only want the voltages. C13 is essential for good fidelity. It may also be a good idea to temporarily disconnet the NFB or go to low, becasue you would have disturbed NFB stability by disconnecting C13.
Should you now find the voltages different as expected, C13 is leaking and must be replaced. It might be necessary to return Rp3 to 220K - you must judge. We need to get the R12 voltage to 90 - 100V and the V3 anodes to just over 200V or so. While 'under' the PC, also see whether anything else is connected from +B3 to somewhere.
I would again need the voltages +B1, +B2, +B3 and at the cathodes of V3 (over R12) to come to a conclusion.
(Just for interest, if you can check: Where IS that absent cap under the yellow wires going?)
But first, forget about measuring at R11 by connecting a meter directly there. The resistances there are high, and even a 10meg voltmeter load will alter the voltages enough to be misleading. (Always when measuring, one must keep meter-loading in mind. Even the best normal meters have an input resistance, and it acts like a resistance added where you measure. In the case of an LTP such as here, and when one is sure the LTP is working more or less normal, the combined cathode voltage will reflect the preceeding anode voltage plus a few volt cathode-g1, which one can guess. The tied cathodes are a low impedance point, which a meter is unlikely to disturb. You actually have no voltage drop across R11; it is the meter.)
But then, on to +B3. That C13 which we have finally identified; unsolder it! I realise it is difficult since you must get to the 'underside' of the PC. Also disconnect/reconnect of wires. We need to know the +B3 voltage without C13, as explained previously. I hope it will be in the region of 150V and that the voltage over R12 will now go to 90 - 100V or so. Please note: This is not a conditon in which to try and listen; we only want the voltages. C13 is essential for good fidelity. It may also be a good idea to temporarily disconnet the NFB or go to low, becasue you would have disturbed NFB stability by disconnecting C13.
Should you now find the voltages different as expected, C13 is leaking and must be replaced. It might be necessary to return Rp3 to 220K - you must judge. We need to get the R12 voltage to 90 - 100V and the V3 anodes to just over 200V or so. While 'under' the PC, also see whether anything else is connected from +B3 to somewhere.
I would again need the voltages +B1, +B2, +B3 and at the cathodes of V3 (over R12) to come to a conclusion.
(Just for interest, if you can check: Where IS that absent cap under the yellow wires going?)
I would add Ca of 100µF although I did indicate that it makes little difference. A 100V capacitor is rather big; the voltage rating do not need to be more that 16V. (Actually 6V, but I do not like very low voltage caps.)
R5C2 has nothing to do with that; it is impossible for me to tell whether they are required. That will only be apparent when the amplifier is operating and monitored with an oscilloscope. The output transformer will influence that.
I take it that by 'DC conditon' you mean without signal. The amplifier is not necessarily stable without load; it should never be operated like that. Input short should not be necessary; if the attenuator is already there, turn it down.
But we still have not dealt with C13! That is more important the the above matters. You need to reply to my previous post before going on.
R5C2 has nothing to do with that; it is impossible for me to tell whether they are required. That will only be apparent when the amplifier is operating and monitored with an oscilloscope. The output transformer will influence that.
I take it that by 'DC conditon' you mean without signal. The amplifier is not necessarily stable without load; it should never be operated like that. Input short should not be necessary; if the attenuator is already there, turn it down.
But we still have not dealt with C13! That is more important the the above matters. You need to reply to my previous post before going on.
Refering to your earlier earlier post.
I am so sorry if I mislead you with my simple sketch. The 0.93mA is shared by the left and right. So each channel has 0.466mA which jive, therefore can I say it is not the B3 capcitor is at fault.
Now to bring the V1a voltage up to about 85V, I need B3 to be 130V. i.e. 0.45mA flowing in V1. for 2 channels, total 0.9mA flowing through 220k. Now to know what is Rp3, I have to assume V2 is 300V, then Rp3 = (300 - 125)/0.9 = 195k.
From your figures
V3 cathode = 90V
V3a anode = 208V, I = 2.8mA
V3b anode = 202V, I = 2.5mA
Total 5mA x 18k = 90V = V3 cathode and so the figures jive.
To know what is B1, total current flowing = 300V + 5mAx2 channel x 2200 = 312.2V
It looks like I have to change Rp3 to about 180~195k to get the voltage of V1a at 85~90V.
I need some help here.
Referring to #53 circuit,
As I assemble back the amp (I have changed the secondary tapping from 276V to 266V), I did a measurement at zero signal (DC condition)
I measure B1+ as 323V. I found 6v6GT plate voltage is 315V. To my surprise the ultralinear connection screen grid no 2(pin 4) is 321 V.
Why is the screen grid 2 voltage higher than the anode voltage. Can it be a wrong connection ???
Referring to #53 circuit,
As I assemble back the amp (I have changed the secondary tapping from 276V to 266V), I did a measurement at zero signal (DC condition)
I measure B1+ as 323V. I found 6v6GT plate voltage is 315V. To my surprise the ultralinear connection screen grid no 2(pin 4) is 321 V.
Why is the screen grid 2 voltage higher than the anode voltage. Can it be a wrong connection ???
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.