Express 5:25 Bias Mod

[wagdog]

Greetings all,

I've been using my Express 5:25 for just over 10 years. It was a floor model at guitar center when I bought it (they were blowing them out), so after a few months of use, I ordered new MESA brand power tubes from Sweetwater. The first set redplated fairly quickly, and Sweetwater replaced them. I also built a bias probe for EL84s to see what was going on.

What I found was there was close to 400v on the plates, and the coolest of the two tubes was idling at 28ma, dissipating 11.2w - yikes. The other tube was was idling at 35ma. Not a well matched set in my opinion, and certainly running too hot, which explains the redplating. After contacting MESA, they assured me their tubes were meant to run in their amps at those levels, and to make things right, they sent me 4 sets of matched tubes, which I appreciated. I found the tubes they sent also had poor matching, and ran very hot. I ended up contacting Eurotubes, who sent me their coldest JJEL84s. Those were running just above the 70% mark (20ma?).

Anyway, my amp is now way out of warranty, and I've finally got time to add a bias pot to it. Here's what I've come up with:

This replaces the original bias circuit currently in the amp. The R1 and R2 replace R54/R55 from the original circuit.

I'll be pulling R54 and R55, which equate to an 8.9k (approx) resistor. Together with R93, those two resistors form a voltage divider, providing approximately -12v to the grids of the EL84s. As that voltage goes more negative, less current passes through the tube at idle, and as that voltages gets close to 0, more current passes through the tube.

I chose a 10k pot for the bias pot as that's the only value I could source for a slot adjustable pot. I believe the voltage swing between -10.3v and -17.4v should be sufficient so I can use most tubes. If not, I'll have to reassess those values. The reason I'm not going to continue to buy MESA tubes is I'm finding the matching is not very good, and they (in my opinion) are biased too hot. No, I don't think I know more than the obviously well qualified designers at MESA, but they aren't paying my tube bills, so I'm going with what I know. I also have a good stock of other EL84s from other manufacturers and I want to be able to use those.

I'm curious if anyone here has done a similar mod and has any input on this?

I'll update the thread as I proceed.

 

[user 21540]

Wow -- Mesa is known to run a lot of their EL84 amps very hot but maybe there was some component issues (drift or just way off spec) in your bias circuit given that the Eurotubes cold set were running that hot as well. Were you able to check or verfiy the values of any of the stock componenets before you modified it?

As a side note a lot of amp manufacturers seem to want users to burn through EL84s quickly by biasing them way too hot in their designs.

 

[wagdog]

@wildschwein - I haven’t modded the amp yet. I confirmed the values of the three resistors in the bias circuit and they are spot on. What I haven’t done is check the voltages in the bias circuit while the amp is running. I’ll do that before I change anything.

When I spoke with MESA back in 2013 I gave them the numbers I was seeing and they ran them by their engineers. They said something to the effect that some of their EL84 amps do run the “hotter” tubes outside of the recommended max dissipation and it’s normal. They assured me that I should just buy MESA rebranded tubes and all would be fine.

So I guess redplating EL84s is normal?

What sealed the deal for me was the poor matching.

Over the years I’ve looked into this issue and there is quite a bit of chatter online about this issue. I don’t think I’m alone in noticing it.

 

[wagdog]

I uploaded a video demonstrating how overbiased my MESA Express 5.25 is. In the video, I use a set of unopened MESA EL84s I purchased last week. I got a bit long-winded in the video - the actual measurement of the tubes starts about 8:30.

I'm going to shoot a video of installing the mod at some point, maybe next week.

Video link:

Start of actual testing:

 

[wagdog]

Hey all,

In the video below, I explain the bias mod I'm going to do. The next video will have the mod as I performed it.

 

[Nandrewjackson]

I suggest attaching the unused leg of the bias pot in your drawing to the wiper leg. That way, if there's ever anything goofy going on with the pot, it will "fail" towards maximum negative voltage, instead of possibly losing -VDC reference to ground.

 

[wagdog]

I suggest attaching the unused leg of the bias pot in your drawing to the wiper leg. That way, if there's ever anything goofy going on with the pot, it will "fail" towards maximum negative voltage, instead of possibly losing -VDC reference to ground.

Thanks for that, I may do just that.

My thought is if the pot fails (wiper disconnects for example), the bias will receive max negative bias voltage as there won’t be any current going to ground through the pot and 7.5k resistor This would effectively shutdown the output tubes, but I don’t believe it will result in any harm to them.

Thoughts on that?

 

[Nandrewjackson]

Yes, I see your logic, too, and I'm not saying it's wrong .

The way I understand secondary windings is that they need a reference. Like how the ac heaters can be lifted , or how zener on the CT can lower overall Voltage. I'm not an engineer so I wouldn't be surprised if my understanding is wrong. I'd at least put a big juicy 1 Meg parallel to that cap .

 

[wagdog]

Yes, I see your logic, too, and I'm not saying it's wrong .

The way I understand secondary windings is that they need a reference. Like how the ac heaters can be lifted , or how zener on the CT can lower overall Voltage. I'm not an engineer so I wouldn't be surprised if my understanding is wrong. I'd at least put a big juicy 1 Meg parallel to that cap .

Gotcha.

FWIW, I’m not engineer by any stretch either. I have been working on amps for quite a few years though (20?), and haven’t catastrophically modded anything, yet.

My take on the reference to ground for the 55v secondary (bias supply) is that it is through the red/ylw wire, which is the same reference as the primary.

The heaters could be lifted by referencing the brn/ylw wire to a higher DC voltage.

Anyway, I don’t think attaching the bottom of the pot to the wiper will have any adverse effect, and agree with your reply that it’s probably better in the long run.

I did complete the mod in the amp (without the bottom pin on the pot connected), and it works, well, spectacularly. I videoed the process and will post it when I get time to edit the video.

When I get a spare minute I’ll go back in and attach that bottom pin as I think it’s probably the correct way to do it.

 

[Nandrewjackson]

Great, another happy mesa running cool

 

[wagdog]

I went back in and added a wire between the unused bottom leg of the pot and wiper as suggested above.

Although I’m still editing the video (coming soon!), in it, I biased the tubes a little cold. I played it that way for a bit and decided to bring them up a hair hotter. Not nearly as much as they were with the stock circuit, but warmer than my initial setting.

I found the amp felt a little livelier overall with the tubes a little warmer. I usually bias my fender amps on the cold side, but the express seems to sound/feel better a hair warmer.

I do know that adjusting the bias supplies a higher b+ voltage going to the preamp tubes down the line. This might make the preamp play a little “stiffer”?

Anyway, having the ability to adjust the bias and use a wider variety of EL84s makes this mod worthwhile for me.

 

[wagdog]

Here is the last video in the series in which the bias pot is installed. It's long.



This is a schematic of the final version of the mod as well:

 

[DrKev]

Well done and thank you, Wagdog! I may do this myself in the coming months.

 

[wagdog]

Well done and thank you, Wagdog! I may do this myself in the coming months.

I’m very happy with it. I ended up running the current set of EL84s a bit over 70%, but well under where they were at.

I generally run my 6L6 amps a bit cold, but have found el84s do sound best a touch over 70%. Nowhere near where MESA had them set though.

Good luck - be safe!

 

[rolandibus]

Hello, I appreciate your bias circuit and will install it soon. However, I have a question: if I switch the switch to 5 watts "class A", the bias current of the left tube goes up to 90mA, but the right tube only goes to 2mA! That's not normal, is it? The tubes are matched and are new, even if I change the tubes on the left with the right, the left one is always at around 90mA Thank you very much for replying

 

[wagdog]

Hello, I appreciate your bias circuit and will install it soon. However, I have a question: if I switch the switch to 5 watts "class A", the bias current of the left tube goes up to 90mA, but the right tube only goes to 2mA! That's not normal, is it? The tubes are matched and are new, even if I change the tubes on the left with the right, the left one is always at around 90mA Thank you very much for replying

When you switch to 5w mode, the power amp circuit is changed from being a push-pull fixed bias configuration, into a single-ended (SE) cathode biased configuration (sort of, as I would have expected the B+ to start at the opposite end of the OT in SE configuration and it looks like it still starts halfway).

Short answer: only one output tube is being used in the 5w mode.

Looking at the schematic, it looks like the plate voltage changes to 288v in SE mode on the one EL84 that's being used and it's cathode biased via R104, which is a 20ohm 2W resistor. 20ohm seems low to me, but I've only ever built 6V6 and 6L6 based single-ended amps, so I don't know if that's reasonable.

And, I don't understand what the role of C16 and the 750ohm resistor is, which is connected from B+ to the cathode of the single-ended tube. Perhaps this is for tone shaping when in single-ended mode or maybe it raises the voltage of the cathode for some reason? I really don't know.

Anyway, I also don't know if 90ma a reasonable value for a single-ended EL84 in the "5w" configuration, but it sure does sound excessive to me. It would depend on what the actual voltage is on the cathode is to calculate the drop across R104. But again, I really don't know, so take what I'm saying here with a grain of salt.

For the record, I never use the 5w mode as I found it doesn't have any tones I desire. When I need to be quieter, I use the master volume(s). The Express is an one of the few amps I've owned that really doesn't loose much oomph when turned down - it's one of the reasons I like it so much.

Sorry if that reply wasn't real helpful, but that's about all I can offer about what's going on in the SE mode. It's certainly is something I'd like to learn more about myself just to have that knowledge.

 

[rolandibus]

Hi "wagdog",
thank you for the explanation of the 5 watt mode. That already helped me.
So it's better not to work too much with the 5 watt mode because that puts a lot more strain on one tube of the pair than the other tube.
This is exactly what happened with my amp: one EL84 tube (the left one seen from behind, which then works as a triode) is broken because it was loaded much more than the other!
I have now also found out that there is an error in the schematic and that R104 is not a 20ohm 2W resistor but 120ohm!

 

[wagdog]

Hi "wagdog",
thank you for the explanation of the 5 watt mode. That already helped me.
So it's better not to work too much with the 5 watt mode because that puts a lot more strain on one tube of the pair than the other tube.
This is exactly what happened with my amp: one EL84 tube (the left one seen from behind, which then works as a triode) is broken because it was loaded much more than the other!
I have now also found out that there is an error in the schematic and that R104 is not a 20ohm 2W resistor but 120ohm!

I’m admittedly not 100% certain of the SE operation going on in the express, but it would appear that none to very little current is going through the disconnected tube. This would result in abnormal wear on a pair of tubes as the operational tube would wear out faster over time. Both tubes have heater current, but only one is running current from the plate to the cathode.

I don’t know if you’re using MESA branded tubes, but I found their matching to be very poor. Not that it matters in the SE use scenario.

Anyway, as I mentioned in a previous post, the 5w SE mode was of no interest to me tone wise so I never use it.

 

[Stinger22]

Hi "wagdog",
thank you for the explanation of the 5 watt mode. That already helped me.
So it's better not to work too much with the 5 watt mode because that puts a lot more strain on one tube of the pair than the other tube.
This is exactly what happened with my amp: one EL84 tube (the left one seen from behind, which then works as a triode) is broken because it was loaded much more than the other!
I have now also found out that there is an error in the schematic and that R104 is not a 20ohm 2W resistor but 120ohm!

It's imperfectly fine to use the 5 watt mode that is what it is there for and it has a great class A tweedy blues tone that is still plenty loud and can even be miked for a small gig. Don't be put off because only one tube is being used, it's not going to break it, and if you are that worried about it just rotate them once a year.

 

[rolandibus]

Thanks in advance for the answers
but I can't get any further because there is a problem with the voltages on the two EL84s and the bias of
the two EL84s are very different, although both tubes are matched.
What remains to be said is that it is an export amp and is operated with 230 volts
and all voltages are a little higher, as listed in the schematic!
Here are the voltages from A+-B+-C+
A+= 420V - B+= 360V - C+= 260V

And here are the voltages of the two EL84s in normal mode and in SE (5 watt) mode:

NORMAL MODE:
EL84 1
Anode (7) 420 V
Cathode (3) 1.6 mV !
Grid 1 (2) 17 V
Grid 2 (9) 370 V
BIAS CURRENT: 7.1mA !

EL84 2 (the SE tube)
Anode (7) 420 V
Cathode (3) 14.7 mV !
Grid 1 (2) 17 V
Grid 2 (9) 370 V
BIAS CURRENT: 18mA !
-----------------------------
SE MODE (5Watt mode)
EL84 1
Anode (7) 370 V
Cathode (3) 1.2 mV
Grid 1 (2) 16.8V
Grid 2 (9) 160 V
BIAS CURRENT: 5.8mA

EL84 2 (the SE tube)
Anode (7) 160 V
Cathode (3) 11.3 mV
Grid 1 (2) 0.5V
Grid 2 (9) 160 V
BIAS CURRENT: 99mA
--------------------
As you can see, the bias current is very unequal even in normal mode
Both EL84s are matched and I tried again with 2 more matched Duet EL84
but always with the same result.
I also tested the resistors around the EL84s, but they are all OK!

What remains to be said is that it is an export amp and is operated with 230 volts
and all voltages are a little higher than with American amps!
The reverb tank wasn't connected when testing, but that can't be the problem.
Unfortunately I don't know what else I should check? Does somebody has any idea?
Maybe I should start it as a new post, what do you think?