Silverface Bassman 100 power supply issue

[cage771]

I'm working on a Silverface Bassman and ran into a strange issue in the power supply. At each of the red legs, I have 335VAC to ground and 670VAC between the two. After the first diodes, this drops to 115VDC (stays at 115 after the second diode as well). At the end of the rectifier circuit, I'm seeing ~370VDC rather than the ~470VDC I should have. This stays constant with or without a load.

Bias and heater voltages are correct per the schematic.

Has anyone seen this before or have any pointers/thoughts? I'm pretty stumped at this point.

Background on this amp - it was brought to me due to low volume. With everything knob on 10, it was loud, but not 4*6L6 loud. Keeping it clean, it can't compete with a medium hitting drummer. It should be much louder. I measured the above voltages initially, also noticed that the tubes were biased at ~20mv (cold!). So in I went. I did a cap job on it (still had factory caps) and replaced the diodes in the rectifier and bias sections. I have a fresh set of tubes for it but don't want to install them until I get the voltage issue figured out. A proper 3 conductor power cable has been installed and the death cap removed by a previous owner (I was told that someone had done some Blackface mods to it but I'm not seeing it).

I appreciate any and all help on this one!

[Willabe]

Welcome! 

Are you sure it's a 100? Bassman 100 has 6 B+ diodes.

Baseman 135 has 4 diodes. (FWB)

Is it this 1?

https://el34world.com/charts/Schematics/files/Fender/Fender_bassman_100_schem.pdf

Or this 1?

https://el34world.com/charts/Schematics/files/Fender/Fender_bassman_135_schem.pdf

[cage771]

Willabe -

It is the 100 with 6 diodes (3 in series off of each leg). On each leg I have 335VAC going in. After the first diode on either leg, it drops to 115VDC. I continue to have 115VDC after the second diode on either leg. When they come together, I have ~370VDC. These numbers remained constant from before and after I replaced them. I see the ~370VDC not matter if the filter caps are connected or not. I am also seeing the same ~370VDC at the standby switch (as opposed to the 470VDC I should see). This 100 volt difference is strange.

Something that I did not do - replace the resistors in the bias and power filter circuits. I checked them to verify the values (solid per spec) and I don't see how they would impact this particular issue. 

BTW, I am working off of the exact Bassman 100 schematic that you linked to. I've been elbow deep in this chassis since Monday when my parts arrived (took Wednesday night off for band rehearsal). This thing is driving me nuts!

Thanks!!!!! 

[2deaf]

When they come together, I have ~370VDC. These numbers remained constant from before and after I replaced them. I see the ~370VDC not matter if the filter caps are connected or not.

370Vdc whether the first filter stage is connected or not strongly suggests that the first filter stage is not really connected at any time.   

[cage771]

2deaf -

Thanks for the comment and I will verify that connection again. I'm rather meticulous when it comes to cap jobs, and I measure everything before putting the dog house back on, but something might have opened up. What you're saying makes a lot of sense though.

Color me stupid, but why would I see B+ voltages if the first stage filtering cap isn't connected? I'm feeling snowblind at this point due to looking at schematics and this thing for so long.

I do appreciate the "second set of eyes" on this!!!

Brandon

[cage771]

2deaf -

Thanks for the pointer! I was thinking the same thing regarding no load/load but was fried. I opened the dog house and sure enough, the first filter cap wasn't connected. Somehow, the leg was broken between the cap body and the solder joint. I don't have a clue as to how that leg broke through.

Nonetheless, this issue is resolved! I'm around 460VDC at the standby switch now. A little low, but with sew tubes and some adjustments, I can get it on mark.

Thanks to both of you for your thoughts and ideas!!!! Much appreciated! I'm going to start hanging out here more often.....great information.

[2deaf]

... why would I see B+ voltages if the first stage filtering cap isn't connected?

When you connect the meter probe or your finger (don't do that) to the diodes, the circuit is completed and there really is DC at that point with a lot of ripple.  A LOT of ripple.  The meter interprets this pulsating DC as some mysterious DC value that varies from meter to meter, but it frequently seems to be around three-quarters of the peak voltage.  If you set your meter to AC, it usually reads a much lower voltage that is even more mysterious.   

[cage771]

2deaf -

I looked back through everything and can see why I has getting sound but the voltages weren't correct. I went down a rabbit hole and became focused on one thing. Tunnel vision per se'.

I'm now sitting at 457vdc at the standby switch in the "on" position.....13vdc lower than the spec of 470vdc. I'm not worried about that loss as it should be within the +/- range. I do need to figure out why my bias is low. I haven't done the mod, but using the balance pot, I can get one pair to -37mv......but the other pair reads at -12mv. Balanced gets me -25mv.....which is rather cold for a 6L6GC. 70% at 457vdc should be -45vdc or so.

So I will need to do the math to figure out the correct resistor value to get in spec. Fun stuff!!

Thanks again!!!!

Brandon

[shooter]

-25mv

Isn't that kinda close to 0vdc, which is past biased hot, closer to biased melt ?

[2deaf]

Isn't that kinda close to 0vdc, which is past biased hot, closer to biased melt ?

No.

[2deaf]

I haven't done the mod, but using the balance pot, I can get one pair to -37mv......but the other pair reads at -12mv. Balanced gets me -25mv.....which is rather cold for a 6L6GC. 70% at 457vdc should be -45vdc or so.

Which mod?  Do you mean -45mV which would mean 45mA, or an actual bias voltage of -45V?

[cage771]

2deaf -

I was speaking of the bias mod.....making the circuit into fixed. I'd like to do this, but I need to get everything working as normal before making changes.

I miss typed. I should be seeing -45mV or 45mA, Neg 45VDC would be very WAY off! I'm using a bias probe from Eurotubes. Doing the math, my numbers reflect what I am seeing on the meter. I have a dedicated meter for the probe and use a Fluke 87 for everything else. My scope is handy, but not for this.

Thanks again!

Brandon

[2deaf]

-45Vdc for the actual bias voltage is plausible with a cold set of tubes.  I was 98% sure that you meant 45mV, but I get burned over and over again on that 2%.

The schematic calls for 447V with the standby on play.  Once you get some more current going through the output tubes, your B+ will drop down a little.

Once you get there, are you looking to convert the existing 10K pot to a bias level or to add another pot and retain the existing balance configuration?

[cage771]

2deaf -

Thanks for the follow up. I know what you mean!!

I took some additional readings. At the joint between the bias cap and bias diode, I'm seeing -70.6 (spec is -69). Going into the balance pot, im seeing -58 (spec is -57). At the standby switch, Im seeing 471 (spec is (470) when off and 457 (spec is 447) when on. These numbers all seem pretty good, but when I use my bias probe, I'm only seeing 25mA (which would also be -25mV). To get up to 70% dissipation, I should be seeing 37mA or so. I'm not seeing the power that this amp should be capable of. Side note, I'm using an 8 ohm cab. Oh, and I installed a brand new balanced set of JJ 6L6GCs.

Update - I traced my voltages through the resistors in the bias circuit. I have -57 on both sides of the 68K and the 1.5K resistors. This is balanced and hitting all power tubes equally.

Looking at the schematic and thinking out loud, I need to verify the resistance of the 68K and 1.5K resistors going to the grids.

As for the bias mod, I haven't decided which direction I want to go. I also want to convert the master volume to a PPIMV....but not until I figure out the rest of this. As I said, I think that I'm snow blind from looking at this thing for too long.

Thanks!!

Brandon

[sluckey]

I have a fresh set of tubes for it but don't want to install them until I get the voltage issue figured out.

It's time to put those new tubes in.

[cage771]

Slucky -

I put the new tube set in last night to check things out. Same issue. There is something going on that I have yet to find.

Thanks!!!

Brandon

[2deaf]

You need to turn the bias down in order to raise the idle current (down is up for bias today).  There are lots of ways to do this.  If you just want to bias this set of tubes, probably the easiest thing to do is to change the 470r resistor between the winding and the diode.  Changing to 820r drops the bias by about 3V, 1K about 5V, 1.2K about 7V, and 1.5K about 10V.

If you are going to add an additional pot to adjust the bias, it usually replaces the 15K resistor going to ground.  For now you can parallel that 15K with maybe a 47K and see what happens.

[cage771]

2deaf -

Thank you! This is another pointer that I needed. Unlike with other repairs, mods, and builds, I let myself get bogged down by one idea and ignored the basics. Judgement has become clouded if you know what I mean.

Unfortunately (thankfully as well), I have a date with my wife tonight.....whoohoo no kids! I might be able to get back on this project tonight but otherwise I will first thing in the morning after coffee. I have parts so I will have updates soon.

Thanks again!!!

Brandon

[Slimtim]

got out there and have a blast on your date.that old amp aint goin anywhere.

[cage771]

You need to turn the bias down in order to raise the idle current (down is up for bias today).  There are lots of ways to do this.  If you just want to bias this set of tubes, probably the easiest thing to do is to change the 470r resistor between the winding and the diode.  Changing to 820r drops the bias by about 3V, 1K about 5V, 1.2K about 7V, and 1.5K about 10V.

If you are going to add an additional pot to adjust the bias, it usually replaces the 15K resistor going to ground.  For now you can parallel that 15K with maybe a 47K and see what happens.

2deaf -

I actually have some 50KL pots on hand. Two different form factors (surface mount and through hole mount). I need to figure out how I'm going to mount one where I can easily access it. I will replace that 15K with a pot and see if I can get the voltage in the proper range.

Brandon

[2deaf]

A 50K pot. isn't a very good choice in this situation.  I just mentioned temporarily tacking a 47K resistor in there to show you what happens when you decrease the 15K resistance.  Replacing the 15K resistor with a 10K pot in series with a 7.5K resistor works pretty good.  That gives you a range of 7.5K to 17.5K.   That's a range of much hotter than stock to a little colder than stock.  Given that Fenders are usually biased way too cold, there doesn't seem to be much sense in going much further in the cold direction.

If you change the whole bias circuit to a higher impedance in order to use a 50K pot., it drastically reduces the effect of the 10K balance pot. since the 10K pot. would be much smaller relative to the rest of the new circuit.  You could parallel a resistor with the 50K pot. if you really wanted to use it.

[cage771]

2deaf -

I agree, the 50KL pot isn't the best choice in this situation.....but it's what I have on hand on a Sunday (an I'm out of 47K resistors). With the closure of all of the small parts suppliers, I'm using what I have as a temp test. Right now, I have the pot wired in place of the 15K resistor (dialed in at 15K as well). I will adjust downward accordingly as I watch my neg voltage and mA on the tube. I'm thinking that if I can get the plate dissipation to 70% or so, I can check the value of the pot, and install the appropriate resistor for this quad of tubes in place of the 15K. That's my working model anyways. I'm going to pull the power tubes and check voltages after I finish this quick break.

I'm basing my thought process on this - https://robrobinette.com/Silverface_Amp_Mods.htm#Convert_Bias_Balance combined with the numbers and my experience with Marshall JTM and JCM circuits. JTM 45 were pretty much Bassman circuits with a few tweaks.   

I do appreciate your words of wisdom on this issue. Information and condemnation is always welcome. Updates to come!

Brandon

[cage771]

I installed the pot removed the tubes, and verified voltages. I then installed the tubes and checked again. I have ~456VDC at both the Plates and Screens of the 6L6s. Grids are -42VDC and -43VDC after the balance pot. ~457VDC across the standby switch in play mode. Bias meter is reading ~46mA....should be right about 70% plate plate dissipation if the bias calculator is correct.  Anything seem odd?

Thanks -

Brandon

[sluckey]

That looks fine. What about the other three tubes?

[cage771]

Sluckey -

That is a great question. All four output tubes are sitting within a milliamp of each other via the bias probe and calculator. Good matched quad, the balance is right, and I've got the bias voltage humming along nicely.  The amp is much louder now.

Now that I have this figured out and fixed, I have to change the master volume. New post to come on that!

Thanks!!!!

Brandon

That looks fine. What about the other three tubes?