Mesa Dual Rectifier 3-Ch Solo wth LF oscillation from 5U4GB's (SS Mode)

[nevetslab]

I have a client's Mesa Dual Rectifier 3-Ch Solo Amp head on the bench. It came in due to Low Power Output, though I didn't find that to be true (91W Out @ 8 ohms, visible clip), but after listening to the amp and trying to sort out some really nasty noise when the Effects Loop is IN AND the Send Level set to "Normal", I had to step back and set up equal gain on the 3 channels (measured at the Effects Send jack), and got what looked like respectable noise measurements from all 3 channels (-60 to -70dBV wideband lower A-wtd)....Ch 1, having only 3 tube stages was the quietest, with Ch 2 & Ch 3, having 3 additional tube stages were a good 10dB noisier.

When I went to listen to it thru my 4 ohm test speaker (Ampeg BXT-115-L4), it was obnoxiously noisy, not making any sense with the measured data.
And, having earlier been running the amp with just the outer pair of power tubes installed and the two 5U4GB rectifier tubes out, I restored the 4 6L6GC's and the pair of rectifier tubes. Turning down the Effects send got rid of the noise, but also the overall gain. Bypassed the Effects Loop, now quieter, but lower gain. After plugging my bass in and listening to Ch 1, it was just downright strange, no LF definition, distorted, just odd. When I stopped playing, I became aware of a low freq pitch from the speaker, NOT HUM or AC mains related

I unplugged the bass, remaining in Ch 1. Still hearing some very low freq sound. Switched back to Standby, then back on again. Within a few seconds that Low Freq sound was back, like on the verge of motorboating. I switched to the Tube Rectifier mode, and the LF sound quit. Switched back to SS Rectifier, and within a few seconds, it came back. I haven't yet checked to see if it's on all 3 channels, and dependent on the Gain & Master Volume settings. I had to leave to catch my bus back home, but did find a post on this forum from another member having found this problem on a Triple Rectifier Solo head, and after changing the rectifier tubes, the problem went away.

The Cathodes of the two 5U4GB's are connected to the 1N4007 rectifier diodes, all feeding the 1st stacked filter stage ahead of the Standby Switch. Switching between the two rectifier modes is connecting the HV Secondary leads to either the diodes or the Anodes of the tube rectifier's, while the 5VAC winding is connected to the heater/cathodes of the tube rectifiers (and to the first filter stage).

I don't recall coming across this one before, and I only found one thread on it, describing the sound to that of a big truck coming down a nearby alley, hearing that low frequency rumble. I don't yet know if this LF oscillation is also related to the rest of the system noise. I'll resume tomorrow to see what's what.

Didn't find anything obvious on the PCB's, removed the rear panel connector board for inspection, no solder fractures on it or what is exposed fro the open chassis. Reseated ribbon connectors, but haven't tried to tear apart the amp. I did swap the one pair of 5U4GB's with another pair, though not brand new. Still had the LF oscillation.

How does this LF oscillation couple into the amp? Is this a common known problem, or do I have an oddball here?

[nevetslab]

Two versions of the Tube/SS Rectifier selection wiring

Just resuming the search on this LF oscillation that occurs when in Solid State Rectifier mode. When I downloaded the Preliminary schematics of the Dual Rectifier 3-Ch Solo head, that shows the Tube Rectifiers ALWAYS in circuit, with using a SPST switch to connect the two pair of 1N4007 rectifier diodes across the 5U4GB, jacking up the voltage (as the loss thru the 5U4GB can be 40-50V, it being current dependent). The later version switches the secondary windings to either the 1N4007 diodes or the 5U4GB tubes, NOT leaving the tubes always connected (though the heater/cathodes ARE connected always.

I looked to see which version this amp is. Rectifier switch is SPST, so this one ALWAYS has the 5U4GB's in circuit.

boogie_dualrectifier_3ch_solo_head.pdf

Now I'm off to see if this LF oscillation is the root problem of all the hum/noise with the preamp and the nastiness that the Effects Send pot dials up. I think I like the later version, switching the Anodes if this LF Oscillation potential is always lurking in the background.

LF Oscillation fundamental is about 15Hz, with a strong harmonic at 45Hz (what we're hearing). Master Volumes of the 3 channels don't affect it, though the Effects Send level does a little bit. Also discovered there's HF oscillation that sometimes pops up coming out of the Effects Send

[nevetslab]

Further digging, removing the two 5U4GB's kills the LF oscillation (expected that). But also found with them installed and in SS Rect Mode, if I selected either Ch 2 or Ch 3, and switched it's voicing switch (RAW, VINTAGE, MODERN) to MODERN, it killed the oscillation. That circuit appears to be tied to the LTPI circuit via relays, connecting a 100nF cap to ground across the bottom tail resistor 4.7k, while at the same time, opens the Feedback path between the LTPI and O/T Secondary 8 ohm tap. (See schematics in previous post).

[nevetslab]

Switching Anodes (5U4GB & 1N4007 diodes) with common cathode still oscillates 15Hz

While pursuing the Ch 1 HF oscillation problems on the Mesa DR 3-Ch Solo amp head, the DPDT toggle switch arrived, allowing me to replace the SPST toggle used to switch in the solid state rectifier string across the tube rectifiers (as was done on the 1st generation Dual Rectifier amp. I'd already found that with the 5U4GB's installed, and selecting SS Rectifier Mode, the amp will produce & sustain a 15Hz oscillation, with an amplitude of about 3.5V RMS, or 3W into 4 ohms. What you actually hear is the 3rd harmonic of that (45Hz), which is about 24dB down.

Well........while I was apprehensive about leaving the heater/cathode connection tied in with the 1N4007 diode strings feeding the series stacked 220uF/300V 1st stage filter (with 220k ballast resistors across each), I went ahead to try it. Didn't make one bit of difference.....it still oscillates. Sounds like a big truck coming down a back alley rumbling along.



At idle, 120VAC AC mains applied, BOLD Mode, with SS Rectifier Mode selected the 1st stage Power Supply filter sits at 468VDC, while switching to Tube Rectifier Mode, it drops down to 445VDC (and the LF oscillation quits).

This LF oscillation runs while in Ch 1, Ch 2 or Ch 3 preamp channel, but if in Ch 2 or Ch 3 and the Voicing switch is set to MODERN, that kills it. In a later version of this amp, where each channel has a 50W/100W switch, selecting 50W also kills the LF oscillation.

The effect I'm hearing is a dissonant note modulating what you're playing, and it just sounds awful. And, I still have this Ch 1 HF oscillation issue that can be dialed in with the preamp controls. Sometimes that has been audible oscillation, but normally is around or above 20khz. Another modulation source to contend with, and it has audible effects.....adjusting any of the controls that start/stop that HF oscillation will cause a abrupt LF Thump, then when oscillating, the controls sound scratchy, as though dirty or have DC on the pot.

I haven't yet set up the spectrum analyzer to monitor the full bandwidth (out to 20khz or 50kHz) to see if that HF oscillation is running while playing thru the amp.

Next thought to try is to leave the anodes attached, and switch the rectifier's cathodes to the 1st Filter Cap stage. I still don't understand the mechanism of the oscillation, though I'd bet someone like R.G would.

[nevetslab]

Why does opening the feedback loop via Relay RY7b kill the oscillation?

After no success with switching the PT HV Secondary to either the Tube or SS rectifier's Anodes, and the cathodes remain connected to the 1st power supply filter stage, I resisted the temptation of switching the cathodes of those, which would no doubt emit a substantial pop every time you switch it, I can't stop wondering why opening up the feedback loop kills the LF oscillation. Ch 2 & Ch 3's Voicing switch, when either of those channels are selected, is set to MODERN, it opens the 47k resistor & series 100nF cap from the 8 ohm secondary tap of the O/T to the feedback side of the LTP. Why is there an additional 100nF cap in series, as there already is one between the tail circuit and the bootstrap resistor feeding V5A's grid? Looking at the other end of the FB' line where it ties into the Ch 2 & Ch 3 Presence controls via Relays RL8 and RL9, there are again series 100nF caps ahead of the N.O. relay contacts, though those are part of the Presence control circuits.

boogie_dualrectifier_3ch_solo_head.pdf

I tried to find that 100nF cap that, on the schematic, is connecting to the 8 ohm secondary tap. Found the 47k resistor, and the 4.7k shunt to ground on the tail side of the LTP circuit, but can't find that 100nF. I wanted to at least see if shorting that cap out changes or kills the LF oscillation. Not sure why it is in circuit.....not there for DC blocking purposes. The 100nF caps ahead of the Presence control are part of that circuit.

But, for the time being, I've lost so much time chasing my tail and a LF oscillation condition that appears to be a permanent attribute to the Dual Rectifier 3Ch Solo amp circuit (along with it's Ch 1 HF oscillation), that I just packed the bloody thing up and put it back into the road case, before unwillingness to let go leaves me with no income to show for December. I was hoping for some help out there from those who know these amps. Unplugging the 5U4GB for certain kills the LF oscillation. It still sounds like crap even without the LF oscillation. So for the time being, I admit defeat, and move on to making a living again.

[Jazz P Bass]

Whoo.
Can't say you didn't try.

[TimmyP1955]

Write it off as a noble failure, and tell the customer to send it to Mesa.

[Pedro Vecino]

The only oscilations I've seen in Dual rectifiers occurred installing certain preamp tubes in V1. Appears in pushed mode from the first channel when gain and treble are adjusted near to maximum. This tendency to oscillation is variable between units (not identical) and the amplifier has some components that are intended to limit, such as 1n capacitor between anode of V1a and ground. 250p capacitor through the clean channel master also colaborates to it.
I have also seen a low frequency hum sometimes when the fet through the 475K resistor previous to V2a grid is shorted and has allowed to pass DC towards the grid. In presence of hum, it's best to disconnect the fet to check it.
It´s possible that the association of your problem with rectifier tubes is purely circumstantial. However these voltages you mention do not fit at all with European Rectifiers (230V version). In bold mode, with diodes usually are 450V and with 5U4 down to 415 approx.
The dual-off switch used to activate tube rectifiers in the Dual first version is less practical than the latter through a SPST. In fact I have converted many to the latter system to use finally a DPDT and automate a bias readjustment when using tube rectifiers. I never appreciated problems with this system.

[nevetslab]

LF Oscillation ONLY OCCURS in 4 ohm connections

Having finished moving from Gardena to Glendale, CA, and flushed out all the other work in the shop, this Mesa Dual Rectifier 3-Ch Solo Amp is still here, waiting for me to be brave enough to tangle with it.

Yesterday, I set it up on the bench, out of the case, guts side up, sitting on the lab jacks. With the amp connected to my 8 ohm resistive load, I never saw it produce the LF oscillation. I read thru my notes from December, finding it would start up on it's own, after coming out of Standby, even with the Output control turned down (overall Master Vol).

I fired it up again this morning, now with the 4 ohm Ampeg BXT-115HL4 bass cabinet connected (20 ft away), connected to the 4 ohm output jack. Came out of standby, and saw the LF oscillation start up. Turned down the Output level control, made no difference. Solid low level 13Hz oscillation (had been saying it was 15Hz, but, who's counting).

I jacked the speaker into the 8 ohm output, and NO LF Oscillation. !!!???? Changed to my 8 ohm cabinet, still NO LF Oscillation. Connected that speaker back to the 4 ohm output....LF Oscillation again. I hadn't noticed that before.

I had also made the comment about the 100nF coupling cap between the Feedback tap of the O/T and the feedback circuit connected via LDR circuits, used to open or close the loop. There IS DC Voltage there, off the cathode circuit of the LTPI circuit, though most other amps we see never bother adding a DC blocking cap at the O/T Secondary side of the feedback resistor. I still never did find that cap, but don't think it's related here.

I haven't verified the statement, but recall reading in another thread that CH 1 operates with NO Feedback. Ch 2 & 3, when the voicing switch is set to 'MODERN', kills the LF oscillation (as it opens the loop). I'm thinking CH 1 ALWAYS has feedback connected, since opening it kills the LF oscillation in Ch 2 & 3.

Still don't know the cause, but.....it is a significant finding. Or,, just interesting, if nothing else. Everyone should have a bit of LF modulation present to make them feel a little seasick while playing.

[Jazz P Bass]

Keep chipping away at it.

Too bad it did not 'fall off of the moving van'.

[nevetslab]

Short cables = NO LF OSCILLATION

More fuel for the fire. My test speakers are at the end of the long 33' x 10' shop I have here (used to be a cluster of closets, that later had the walls removed for a larger odd-shaped storage room. So, my test bench being at around mid-point in the room, I've lived with a 20 ft cable to attach my test speaker to the subjects under scrutiny.

What happens under 'normal' conditions, amp atop local speaker, short 2-3 foot cable? Hah! NO LF OSCILLATION!! Plugged it into an old Crate amp that's been holding up coffee cups, spray cans and such next to my chair. Connected it to it's 8 ohm 12" speaker, and all I get from it is the other problem.....HF oscillation. Though that you have to get radical with the controls cranked to nearly max before it comes unglued....not normal settings. It was a lot easier to get the HF oscillation out of it with the 20 ft cable. So, adding additional LCR to the speaker and you get more of a tuned circuit, which this Mesa apparently likes for LF sustained oscillation.

It also WILL NOT oscillate in resistive loads, even with a 20 ft cable attached. Another reason they're called 'dummy loads'!

I may get this amp out of the shop yet. It now seems to be left with the criteria one of Mesa's staff commented about: Richard Duvall @ Mesa puts it like this..."We design the amps to have as much gain as the customer can get in trouble with"

[nevetslab]

uh.....make that no SUSTAINED LF OSCILLATION. After putting the amp back into the case, moving the speaker stack close enough to reach the amp with the 30" cable, and being a bass player, not a guitar player, playing down in first position on the neck with the 4 ohm Ampeg cabinet connected to 4 ohm output jack, I was getting ugly sounding LF modulation. Patched in a long cable to reach the analyzer/scope, and saw what I didn't wanna see.....that bloody LF modulation. Changed to the 8 ohm speaker and jack, and it sounded just fine. Back to the 4 ohm cabinet and jack, Tube Rectifier mode, no problem. Back to SS mode, and the spongy regulation, no problem. And playing up higher on the neck, I wasn't having any problem, so......I think I've gone as far as I care to, short of pushing it out of a moving van!! I did remove gain in a couple locations to reduce the tendency to oscillate in the HF region. I didn't get it oscillating with settings that could still rip your ears off, so, time to move on, and see if this sounds acceptable to the client at this point.

[alansom]

mesa boogie triple

hello good afternoon

I bought a boogie triple and it came with the faint sound, but very weak
then I switched valves and it remains weak
anyone seen this?

[Jazz P Bass]

Can't say that I haven't seen this.

At any rate, the first thing to check is the power supply.

If the opamp power supplies are good & all of the tubes have there proper voltages (plate & grid), then it is time for a 'send a signal through' test.

[Pedro Vecino]

Firstly select Bold mode and "diodes" in rectifier section. Turn loop selector to bypass (this completely removes loop, general master and Solo function). Try again.