Selmer Concord 10 Rebuild/Conversion Issues

[BungleSim]

Here are the voltages I measured tonight. This is with nothing plugged into the input and the volume at 0. Remember that the tremolo is not connected (see previous posts).

B+0 = 327.5V
B+1 = 306.4V
B+2 = 240.7V
B+3 = 163.2V

V1 (12AX7 preamp)
1 = 108.0V
2 = 0V
3 = 0.908V
6 = 103.8V
7 = 0V
8 = 0.878V

V2 (12AX7 inverter/tremolo)
1 = 182.6V
2 = 38.8V
3 = 55.2V
6 = 101.4V
7 = 0.011V
8 = 0.944V

V3 & V4 (EL84s)
2 = 0.011V
3 = 10.36V
7 = 324.5V
9 = 306.4V

DC resistance of OT primary, between pins 1 and 2 = 75.0 ohms
DC resistance of OT primary, between pins 2 and 3 = 92.3 ohms

Shouldn't these be equal?! I took these measurements in and out of circuit and they were the same.

Attached is a photo of the power tubes after being on for about five minutes. I guess you could say they're cooking.

[hans-jörg]

Hello,
did you checked heater voltage too?

May be you could try the "light bulb test for first switching on" (or so), to check if you really pull so much current. In this case your lamp should be quite bright. I think here is somewhere a sticky about this bulb test.

Hans-Jörg

[BungleSim]

Hello,
did you checked heater voltage too?

May be you could try the "light bulb test for first switching on" (or so), to check if you really pull so much current. In this case your lamp should be quite bright. I think here is somewhere a sticky about this bulb test.

Hans-Jörg

The heater voltages are good. Please see previous posts.

Edit: Sorry. I'm cranky in the morning. I appreciate everyone's input.

[Phil_S]

DCR on the transformer looks reasonable. The number of turns on each side are equal. The outer winding is a longer wire compared to the inner wire. As I'm sure you know, the length of the wire determines DCR. There is some dark art in winding audio transformers. I'm reasonably certain MM would have interleaved the windings (at greater production expense) to keep the DCR symmetrical if that was deemed important. I think we can assume you received a high quality item that is perfectly suitable for the intended use.

I realize now what you were reporting when you said 88mA at idle. That's for 2 tubes. Each tube is half that or 44mA. One thing you might do is check each tube individually, but 44mA is appropriate for an EL84. Collateral verification done two ways from your voltage readings and the schematic:
1) Primary DCR = 75+92 = 167. Vdrop from B+0 to EL84 plats is 3V. 3/167 = 90mA. That's 45mA per tube.
2) Vk = 10.36; R5 = 125; 10.36/125 = 83mA; divide by 2 = 41.44mA per tube.

Static plate dissipation = Va-Vk = 324.5-10.36=314.14; 314*41.44 = 13W. As I noted earlier, this is consistent with expectations for this amp.

I'd note that RC-30 says the max for a pair of EL84 is 17W.

You might like this calculator (use cathode resistor drop method) where he adjusts for cathode current before calculating static plate dissipation, so it drops to 12.3W. http://robrobinette.com/Tube_Bias_Calculator.htm

For a cathode biased amp, you shoot for 100% as it is self limiting (see below.)

Your picture shows the power tubes warmed up and not red plating. I think you have the circuit operating reasonably well.

For some reason yet to be determined, there is a lot of sag. I am too lazy to read back, but I think you said you saw 166mA when driven hard. This is really half that amount as it is per tube (and that is the average per tube.) I am having considerable difficulty imagining that a tube idling at 42mA or 45mA is all of a sudden seeing 2x that no matter how hard you drive it. Either we'd need to see considerable sag in plate voltage to compensate for the rise in current, or we are missing a piece of the puzzle.

If 30V is the sag, I might be inclined to say that's that and call it a day. However, there is the nagging problem is your report of over 80mA per tube. I'm thinking this has to be addressed. Unfortunately, I'm about out of my depth now and nothing looks obvious to me to suggest what would do that. I am hoping someone else will weigh in here.

Aside from thinking I'd prefer higher plate voltage on V1 to give the preamp more clean headroom, I believe all your voltage readings are reasonable. 325V is a little high for EL84's but you are using JJ's which are thought to be stout enough to handle that, so I wouldn't be concerned.

If you read up on cathode bias, it is often described as self-bias. As you drive the power section harder, the common cathode resistor acts as a limiting device by continuously adjusting the operating parameters of the tube. If you really want the expert explanation, Aiken is the guy, about 65% down the page for cathode bias discussion: http://www.aikenamps.com/index.php/the- ... on-biasing

[BungleSim]

DCR on the transformer looks reasonable. The number of turns on each side are equal. The outer winding is a longer wire compared to the inner wire. As I'm sure you know, the length of the wire determines DCR. There is some dark art in winding audio transformers. I'm reasonably certain MM would have interleaved the windings (at greater production expense) to keep the DCR symmetrical if that was deemed important. I think we can assume you received a high quality item that is perfectly suitable for the intended use.

The OT is actually the stock transformer. The PT is the new MM. Just wanted to say this for completeness of information.

I realize now what you were reporting when you said 88mA at idle. That's for 2 tubes. Each tube is half that or 44mA. One thing you might do is check each tube individually, but 44mA is appropriate for an EL84. Collateral verification done two ways from your voltage readings and the schematic:
1) Primary DCR = 75+92 = 167. Vdrop from B+0 to EL84 plats is 3V. 3/167 = 90mA. That's 45mA per tube.
2) Vk = 10.36; R5 = 125; 10.36/125 = 83mA; divide by 2 = 41.44mA per tube.

Static plate dissipation = Va-Vk = 324.5-10.36=314.14; 314*41.44 = 13W. As I noted earlier, this is consistent with expectations for this amp.

I'd note that RC-30 says the max for a pair of EL84 is 17W.

You might like this calculator (use cathode resistor drop method) where he adjusts for cathode current before calculating static plate dissipation, so it drops to 12.3W. http://robrobinette.com/Tube_Bias_Calculator.htm

For a cathode biased amp, you shoot for 100% as it is self limiting (see below.)

Your picture shows the power tubes warmed up and not red plating. I think you have the circuit operating reasonably well.

Thank you for these calculations and the calculator.

For some reason yet to be determined, there is a lot of sag. I am too lazy to read back, but I think you said you saw 166mA when driven hard. This is really half that amount as it is per tube (and that is the average per tube.) I am having considerable difficulty imagining that a tube idling at 42mA or 45mA is all of a sudden seeing 2x that no matter how hard you drive it. Either we'd need to see considerable sag in plate voltage to compensate for the rise in current, or we are missing a piece of the puzzle.

If 30V is the sag, I might be inclined to say that's that and call it a day. However, there is the nagging problem is your report of over 80mA per tube. I'm thinking this has to be addressed. Unfortunately, I'm about out of my depth now and nothing looks obvious to me to suggest what would do that. I am hoping someone else will weigh in here.

I think we are missing a piece of the puzzle because, despite the math, the amp sounds absolutely terrible and I would consider it not functioning at the moment. For the first 25% or so of the Volume pot's travel, the volume and tone is reasonably good, but as you increase the volume the tone just goes to hell and the volume actually decreases. Factor in all of the sag that happens whenever you play anything and the amp is unusable... Also, if the PT can supply 150mA and the amp is pulling that much if not more at times, isn't the amp being driven too hard?

Aside from thinking I'd prefer higher plate voltage on V1 to give the preamp more clean headroom, I believe all your voltage readings are reasonable. 325V is a little high for EL84's but you are using JJ's which are thought to be stout enough to handle that, so I wouldn't be concerned.

If you read up on cathode bias, it is often described as self-bias. As you drive the power section harder, the common cathode resistor acts as a limiting device by continuously adjusting the operating parameters of the tube. If you really want the expert explanation, Aiken is the guy, about 65% down the page for cathode bias discussion: http://www.aikenamps.com/index.php/the- ... on-biasing

I agree that I'd like to see a higher plate voltage for V1. I was planning on playing with that once I actually had the amp working properly. As for supplying the EL84s with 325V, I could just drop that by increasing R22, no?

There's also been talk of putting an EZ81 in so it can handle the current better. This will also increase the B+, which may not be desirable for this amp, right?

[BungleSim]

I realize now what you were reporting when you said 88mA at idle. That's for 2 tubes. Each tube is half that or 44mA. One thing you might do is check each tube individually, but 44mA is appropriate for an EL84. Collateral verification done two ways from your voltage readings and the schematic:
1) Primary DCR = 75+92 = 167. Vdrop from B+0 to EL84 plats is 3V. 3/167 = 90mA. That's 45mA per tube.
2) Vk = 10.36; R5 = 125; 10.36/125 = 83mA; divide by 2 = 41.44mA per tube.

Wait a minute. 41.44mA per tube seems quite high. Shouldn't it be more like 26mA per tube? http://www.ax84.com/biascalc.html?Bvolt ... ction=calc

So I'd want to make R5 about 200 ohms to achieve 26mA per tube, or 52mA total. Pulling ~84mA total for the tubes seems to explain a lot about how it is operating right now, no?

[Phil_S]

Wait a minute. 41.44mA per tube seems quite high. Shouldn't it be more like 26mA per tube?

No. That calculator is for a fixed bias amp. This is a cathode biased amp. Your amp is biased just right the way it is. Look at the way I did the math.

26mA follows the 70% rule of thumb for fixed bias. Roughly: 36mA max @ 70% = 25mA, where 36mA is half the zero signal plate current from the data sheet. This will vary somewhat with plate voltage as there is an inverse relationship

Look at the tube data sheet: http://tubedata.itchurch.org/sheets/191/6/6BN11.pdf
It says to use a single 130 ohm resistor for 2 tubes, zero signal current is 72mA (for 2) and max signal is 92mA or 46mA per tube. Look at the Class A single tube info, where it says zero signal is 48mA, essentially the same thing.

I sense you may be having trouble wrapping your mind around this. Take a bit of time to contemplate it all. Soon enough it will make more sense. Meanwhile, please take my word for it. I'm echoing what I've learned from others on the topic.

[BungleSim]

Wait a minute. 41.44mA per tube seems quite high. Shouldn't it be more like 26mA per tube?

No. That calculator is for a fixed bias amp. This is a cathode biased amp. Your amp is biased just right the way it is. Look at the way I did the math.

26mA follows the 70% rule of thumb for fixed bias. Roughly: 36mA max @ 70% = 25mA, where 36mA is half the zero signal plate current from the data sheet. This will vary somewhat with plate voltage as there is an inverse relationship

Look at the tube data sheet: http://tubedata.itchurch.org/sheets/191/6/6BN11.pdf
It says to use a single 130 ohm resistor for 2 tubes, zero signal current is 72mA (for 2) and max signal is 92mA or 46mA per tube. Look at the Class A single tube info, where it says zero signal is 48mA, essentially the same thing.

I sense you may be having trouble wrapping your mind around this. Take a bit of time to contemplate it all. Soon enough it will make more sense. Meanwhile, please take my word for it. I'm echoing what I've learned from others on the topic.

Haha. I'm an idiot. I was reading up on cathode versus fixed bias and I totally swapped which one was which in my head and then made my post. Yes, it is cathode biased.

I guess, however, that after all of this discourse I still don't know how to attack the overall problem of the amp running hot and sagging so much.

[hans-jörg]

Hi,
do you have the possibility to swap all caps of PS? Or at least to measure them, same the resistors.
I think your problem is there around.


Your amp is not running hot (plates looking good and normal), you are at idle.Your dissipation for class A is top. Maybe some problems with measuring. By math you are at idle.

Hans-Jörg

PS: I still have a problem with your C chain, 20,40,40,40 (that´s to much at the end of chain). Similar amps have there 10µF for V1. The first cap (reservoir) is mostly the biggest (see Marshall, Fender, TW, Dumble, etc.)
I would try 40 (is way below 50 for tube rect.),20,20,10µF, and listen than again. As I understand your problem is with the sound.

If you change to diode rect. you will have B+ with about 30VDC higher

Just my 2 ct

[Phil_S]

I still have a problem with your C chain, 20,40,40,40 (that´s to much at the end of chain). Similar amps have there 10µF for V1.

I don't have technical knowledge on this, but it makes a lot of sense and I'm glad you pointed it out.

Maybe the Princeton 6G2 is similar amp to look at as it has tremolo and cathodyne inverter. It supplies tremolo B+ from the same node as the OT CT. It injects the tremolo wiggle at the cathode of the 2nd gain stage. Fender used 3x 30uf.

As I said, I don't know much, 3x 30uf is more filtering than I expected to find on the 6G2. Maybe the OP's amp will benefit from moving the 20uf to the end of the line, but I'm not seeing where it is causing the reported problem.

I don't recall earlier mention the amp sounds bad, but it may be somewhere in this thread. I'm thinking multiple problems. It is a simple thing to raise preamp voltage and along with it get a lot more clean headroom. Cleaning up the front end signal might be a good idea at this point.

I'm looking at 5E3 voltages as a guide, which I think runs V1 and V2 at around 200V. In this Selmer, Vdrop between B+1 and B+2 is 66V over 22K. That tells me V1 and V2 are drawing 3mA total. The next drop is 78V over 68K = 1.1mA on V1. Using the mA as a constant (it's not) I'll guess the dropping resistors can be something like 15K in place of 22K and maybe 10K in place of 68K.

Assuming there isn't anything actually "wrong" with the circuit, it couldn't hurt to look at the actual design as the problem.

[SilverFox]

First off I don't have a clue as to what is wrong here other than possibly the first resistor was too small and draining the B+. That was addressed.

I did notice there are only 3 but really only two individuals trying to resolve this issue.

Would someone that can evaluate the design like to add something to the conversation? The design looks familiar but I don't have enough experience to know.

I was told long ago in a programming class: If you can't solve the problem in 5 mins, ask some one for assistance.

Now I do know 5 mins is too short to try to solve an amp problem but I think someone with more experience would be of great value here-

No disrespect intended.

Silverfox.

[ToneMerc]

I'm looking at 5E3 voltages as a guide, which I think runs V1 and V2 at around 200V. In this Selmer, Vdrop between B+1 and B+2 is 66V over 22K. That tells me V1 and V2 are drawing 3mA total. The next drop is 78V over 68K = 1.1mA on V1. Using the mA as a constant (it's not) I'll guess the dropping resistors can be something like 15K in place of 22K and maybe 10K in place of 68K.

Assuming there isn't anything actually "wrong" with the circuit, it couldn't hurt to look at the actual design as the problem.

I agree and I think the low R at the front of the PS is fine and the rail cap value string while not conventional should not be a problem. What really stands out is the voltages at V1 plates and cathodes at idle, why such a high value dropping resistor at the B3 node is beyond me. This is not a hot input greasy amp to be used with a harp, nor is it LTP phase inverter. I would drop that 68k to 1.5-4.7k ish for starters and revisit this whole exercise.

My 2 cents

TM

[hans-jörg]

I would drop that 68k to 1.5-4.7k ish for starters and revisit this whole exercise.

TM

I agree, the 68k are way too high at this poit.

I would also disconnect the whole tremolo circuit, not only the wiper from depth.
anode, grid and cathode from the tube pins away.

The amp should work without any tremolo part connected to tube.
As the circuit is satisfying (sound, voltage, etc.) you could hang it in again. If than problems occur, you know where to look.
As I understand the situation of the amp for now, is that the trem is only disconnected from output, but the input is still influencing the whole circuit.
The reason: I think your trem circuit still blows into (pre)amp (by cross talks or current swell - excuse my non profi expression, but I think you understand).

I build until now only one tremolo circuit into my amps, so I could be wrong, but I had some swell problems too although the trem circuit was broken at output side. It was cross talks from "disconnected" tremolo.

Hans-Jörg

[BungleSim]

Alright, so the changes I'm going to make next are reducing R23 and R24 to increase B+2 and B+3, respectively. I will also change C20, C21, C22, C23 to 40uF, 40uF, 20uF, 20uF or 40uF, 40uF, 20uF, 10uF, respectively. I am also considering raising R22 slightly to decrease B+1 a bit (Yes, I know this will also decrease B+2 and B+3 accordingly) and increasing R21. To me, the EL84s seem to really be working overtime and I'd like to reduce their current draw a bit.

As for the preamp's tone, unless the customer asks me to, I don't want to change anything. He likes how it sounded (before the last tech's attempted conversion) so it should stay as is. The sound right now is terrible due to all of the power issues and the sagging.

Another thing to consider is getting a beefier transformer that can put out more than 150mA for the HT. Running this thing right at the limit of what the PT can handle does not seem wise to me.

I have other (less demanding) amps to attend to and I am putting the parts order in on Monday so there probably won't be an update for another week or so. Thank you to everyone for your input so far.

[hans-jörg]

Hello,

I´m quiet sure you don't need a higher mA rating with the PT.
120 mA for 2 EL84, 1 EZ 81, 2 ecc83 is the common rating for PT. With 150 mA you are more than in game, which don´t hurts

Best,
Hans-Jörg