Markbass Little Mark II

[nickb]

If the amp works with the relay bypassed, It's not likely that a MOSFET failure is the cause of the DC offset. As the Dude said, measure the DC on the output with the relay bypassed,ore better remove that link and check on pin 1 of the relay which s the signal OUT from the MOSFETS. Also measure the voltage across the relay coil. Do not connect a load until you resolve the issue, especially with the relay bypassed. One we have that info we can suggest the next step.

One caveat, I've seen these with one slightly low gain FET which can be tricky to diagnose. You need a dummy load, signal generator and a scope, run to full output and look for asymmetrical clipping. The symptom is the the amp cuts out, but only at high volume, as the average voltage on the output is not close to zero so the power supply is shut down and the error LED on the board lights up.

Lastly, I think my suggested figure of 3.3mV is too high for idle current above (post #97) is too high. 1mV is more like it.

[Enzo]

And just because the relay doesn't turn on, doesn't mean the relay is bad. There are circuits that control teh relay, and they may not be working. FInd out before just replacing a bunch of stuff.

[Sowhat]

If the amp works with the relay bypassed, It's not likely that a MOSFET failure is the cause of the DC offset.

I don't know if that's completely accurate, I've had amps in the shop with alarming DC offsets due to MOSFET fatigue that still produced sound, surprisingly not that bad sounding although you could see it plain as day on the scope.

[J M Fahey]

We need confirmation about presence or not of dc offset; so far we are only guessing it, so please measure and post.
Otherwise we run in circles.

[nickb]

I don't know if that's completely accurate, I've had amps in the shop with alarming DC offsets due to MOSFET fatigue that still produced sound, surprisingly not that bad sounding although you could see it plain as day on the scope.

Imagine that all three high side FETS had failed. The negative feedback will try correct the error and set the DC offset back to zero. The FETs are biased almost to cutoff so a small current will flow through the feedback resistor. That will reduce the gate to source voltage and push the output slightly negative. Again the feedback will compensate and raise the output voltage. It won't quite make it to zero because of the dual DC feedback paths but it will be pretty small, well under 0.5V I would guess.

It may be that you are confusing DC offset with average DC level. DC offset is what you get with no signal. Average DC is what you get with a signal. If a FET is weak and if the output current is high enough you will get asymmetrical clipping. In that case the average DC level will not be zero. This is what I was talking about in the caveat above.

[J M Fahey]

Modern SS amps (anything from the 70´s on) have unity gain at DC and MASSIVE NFB, both DC and AC, think 60dB or more, so even with wildly unmatched power transistors they will be whipped in line, so typical 50mV offset or less.

The ones you need to match are the input differential pair ones, because they rule the show.

[Leo_Gnardo]

The ones you need to match are the input differential pair ones, because they rule the show.

Though it is a nice touch to match the outputs. Back when the standard amp at Clair Bro's was the Phase Linear 700-II (imagine that Vern, Flame Linears being used as serious pro PA amps, golly!) they used to remove ALL the output xtors, throw 'em in a barrel, test each one, then remount matched sets. Along with a couple other mods... which I'm sure included matching the input diff pairs.

[J M Fahey]

Yes, of course.
Matching power transistors is vital so they share the load equally, don´t overheat and hog all the current/dissipation to themselves, etc.

I was talking about idle DC offset only.

Adding something I forgot, at least half the offset we usually measure isn´t even because of Vbe difference in input transistors, but in not too exact matching of voltage present at the + and - inputs , or discarding as "unimportant" voltage drop across input resistor caused by input transistor base current.

IF those details are taken care of, DC offset can be as low as 4 or 5 mV ... sometimes even less.

[nickb]

Modern SS amps (anything from the 70´s on) have unity gain at DC and MASSIVE NFB, both DC and AC, think 60dB or more, so even with wildly unmatched power transistors they will be whipped in line, so typical 50mV offset or less.

The ones you need to match are the input differential pair ones, because they rule the show.

This one is a little different as it has two DC feedback paths. That reason alone stops the DC offset from getting as low as we are used to seeing under one set of FETS removed conditon. To complicate things the value of one of the resistors in these paths is missing from the schematic. For these reasons I gave a (conservative) < 0.5V.

[leonv]

I have Combo Head 2. Power Supply is broken. I read the schematic and I'm appalled. No regulation on the secondary voltage? No protection? The high-frequency rectifier is pathetic. One way Low frekvency rectifier to. Preamp with tl074 ? Really? The passive crossover is worse than in the JBL JRX100. After reading the fibers there is no product reliability. An authorized service provider does not provide a schematic and is unable to repair it. Just changing whole boards like monkeys. Markbass talks about quality where you look. I feel sick about the lies they write. What is the probability that the Power also destroyed the Power transistors?
The price of the product is not more than $ 150

[Leo_Gnardo]

Just changing whole boards like monkeys.

Welcome to the brave (?) new world of amp repair. That's how it's done now. Unless you have some super dedicated microscope-eyed tech that enjoys changing out flea size components, and fixing burnt PC board lands thinner than a hair.

Markbass talks about quality where you look. I feel sick about the lies they write.

They all do that, it's called marketing. When they grow up they can be politicians.

The price of the product is not more than $ 150

Do you mean cost of manufacturing? No matter which company, what product, or where it's made, there's typically a ratio of 6:1 or more between list price and cost to build. The higher the ratio, the more profitable.

[nosaj]

Welcome to the brave (?) new world of amp repair. That's how it's done now. Unless you have some super dedicated microscope-eyed tech that enjoys changing out flea size components, and fixing burnt PC board lands thinner than a hair.



They all do that, it's called marketing. When they grow up they can be politicians.



Do you mean cost of manufacturing? No matter which company, what product, or where it's made, there's typically a ratio of 6:1 or more between list price and cost to build. The higher the ratio, the more profitable.

Manufacturing, employees insurance, utilities. Funny thing is the people who complain have no concept of what really goes into making an amplifier. Yeah maybe a 150 worth of parts but if I gave you a 150 dollars worth of parts in a bag I bet it would maybe jingle. Knowledge is power and for some knowledge = money.

Me I yearn for knowledge cause I have no money
nosaj

[Leo_Gnardo]

Manufacturing, employees insurance, utilities. Funny thing is the people who complain have no concept of what really goes into making an amplifier. Yeah maybe a 150 worth of parts but if I gave you a 150 dollars worth of parts in a bag I bet it would maybe jingle. Knowledge is power and for some knowledge = money.

Every now & then I get someone who wants me to build them an amp from scratch. Or even better, gut an amp & build a fresh one. They expect to pay for the parts - a couple hundred dollars no more. Or maybe buy a kit from Mojo or some such place. My planning & labor is worth nothing to them. I'm supposed to do it for fun! That's why I don't build custom amps.

For commercially made amps, yeah not only do you have to pay the line workers, there's a butt load of other expenses plus pay the executives, promotion department, artist rep (who gives some of the product away), pension fund, and don't forget taxes taxes taxes. Then transportation and now you're only up to half or maybe 2/3 of the consumer's price - the store has to make a profit too.

Heck these days, I suppose all you need to make a bass amp is a metal box, line cord, SMPS supply, class D amp, the last 2 are pre built "ICE" modules, a couple knobs and some connectors. Piece of cake! No hour upon hour fiddle faddling with cutting holes for a dozen tubes, no wiring dozens of parts.

[nosaj]

Every now & then I get someone who wants me to build them an amp from scratch. Or even better, gut an amp & build a fresh one. They expect to pay for the parts - a couple hundred dollars no more. Or maybe buy a kit from Mojo or some such place. My planning & labor is worth nothing to them. I'm supposed to do it for fun! That's why I don't build custom amps.

For commercially made amps, yeah not only do you have to pay the line workers, there's a butt load of other expenses plus pay the executives, promotion department, artist rep (who gives some of the product away), pension fund, and don't forget taxes taxes taxes. Then transportation and now you're only up to half or maybe 2/3 of the consumer's price - the store has to make a profit too.

Heck these days, I suppose all you need to make a bass amp is a metal box, line cord, SMPS supply, class D amp, the last 2 are pre built "ICE" modules, a couple knobs and some connectors. Piece of cake! No hour upon hour fiddle faddling with cutting holes for a dozen tubes, no wiring dozens of parts.

i think there are 2 kinds of people, those that learn to fix and those that learn to pay.

nosaj

[nickb]

I have Combo Head 2. Power Supply is broken. I read the schematic and I'm appalled. No regulation on the secondary voltage? No protection? The high-frequency rectifier is pathetic. One way Low frekvency rectifier to. Preamp with tl074 ? Really? The passive crossover is worse than in the JBL JRX100. After reading the fibers there is no product reliability. An authorized service provider does not provide a schematic and is unable to repair it. Just changing whole boards like monkeys. Markbass talks about quality where you look. I feel sick about the lies they write. What is the probability that the Power also destroyed the Power transistors?
The price of the product is not more than $ 150

The 80V supply is not regulated. This is hardly unusual for a non hi-fi audio amplifier. Also, given the very low source, winding resistances and leakage inductance I wouldn't be surprised if it out performs most non-switching supplies at a fraction of the cost and weight. Given there's a limiter to stop it from clipping anyway it's not going to matter 99.% of the time.

The rectifiers in the high frequency path BYV32, BYV27 and SF22 and all 35nS or less super / ultrafast types so I don't understand your concern. If they weren't they'd burn up in seconds. TL07x series are almost ubiquitous in guitar amps as they offer incredible performance for next to nothing , negligible input currents, excellent gain bandwidth, negligible distortion with feedback, noise figure is the weakness but it's still fine for what are fairly decent sized signals coming out of a guitar pickup.

There is overvolatge protection on the PSU, but that's it. It's design choice. I have no doubt there was a maximum BOM constraint the engineers had to stay within.

Passive crossover? I only see active equalization. Maybe you mean the tone controls. Well they may be rather ordinary but they do provide a great range of control that people actually use and like.

I think there is much to like and learn from about the overall electronic and mechanical design and that is reflected in the success of the company. If they were unreliable and sounded bad they would not be in business so they did something right.

I do agree, the service centers do seem to be lacking skills. Nevertheless it comes down to economics. It's faster and lower cost to replace the PCB using low paid unskilled people than have an expert work at component level. I bet one person can fix four an hour doing this.