New Amplifier Design - "Murder One" - Submini Pentode, Low Voltage

[panterafanatic]

The HT voltage is about 80 or 90 volts if I remember correctly, the max voltage on the pentode is 100v, so you could get away with anything from 110 or lower pretty safely, possibly higher. That is with 12 volts input, with 9, you'll have lower results I believe. For this you may want a separate 12v 500mA DC power supply.

And yes, it is the same transformer. Though any 22k:8 to 25k:8 should work well.

[max30]

Great update, thank you, 2 questions regaring the latest version, can MAX 1044 can be used still? or new IC is preffered? Could you post the dimensions in mm for new PCB?

[panterafanatic]

I dunno about the dimensions, but you can easily scale the board by printing out a scrap sheet, then scaling accordingly to IC spacing, .1" dunno what it is in mm, for some reason I'm blanking on the ratio atm.

The MAX1044 is preferred imo, unlike a 555 timer charge pump, you can make the frequency far above audio, whereas the 555 takes a little more care to make completely stable for audio use, although it can still be done.

edit: one inch is 24mm come to think of it. so 2.4mm is IC spacing.

[frequencycentral]

I recently asked John Lyons about scaling from DIYLC gifs, he said:

I scale it to a .1 inch grid in photoshop. An ICs pin spacing will match the grid when correctly sized.

The ICL7660S has an identical frequency boost feature to the MAX1044, with the advantage that it's maximum rated input voltage is higher, and it's cheaper. Murder One will still work with a MAX1044, but 12 volts slightly exceeds the MAX's maximum rated input voltage. If you use ICL7660S, be sure to use the 'S' suffix chip, as the standard version lacks the frequency boost feature.

[max30]

in DIY layout creator SW the scaling is 0,1" per row so number of rows and columns shoud give you the exact dimension of PCB.

[Davidgwells]

Just finished building my murder one, thanks Rick!  It sounded great for a little while and now the power supply is dead.  Any ideas why this happened or where to start looking?

[frequencycentral]

I've built quite a number of these. The only issue I've come across is human errors using an AC power supply or reverse polarity DC power supply. Have you measured your DC input voltage? It's probable the charge pump has failed, maybe you have a short somewhere in the diode/cap ladder, or an incorrect wiring? It's worth noting that the caps in the voltage multiplier will hold a charge for many hours unless drained down, any shorts created while tinkering could easily blow the charge pump. I use a ICL7660S nowadays as it's voltage rating is higher than a MAX1044.

[Davidgwells]

Thanks Rick.  I guess I should clarify that it was the wall wart that got fried.  Previous to that, the amp only worked with certain cabinets some of the time and one specific cab all of the time.  At the time it quit the murder one was plugged into a 16ohm cab but powered down.  Moments before that it was working with the cab just fine.  Clearly I have a short somewhere but damned if I can track it down!  I opened up the wall wart and several components are burnt.

[Davidgwells]

What, am I a thread killer??   So I just got a new wall wart, triple checked for polarity, installed a new tc1044 chip an turned it on.  There was a high end whine coming from the speakers but other than that it sounded pretty good until the volume started to fade.  Once again, the wall wart is dead.  I'm certain that polarity is correct and I've traced the signal path countless times and can't find a short.  I'm ready to tear it all down and start over!  Any suggestions anyone?  Why am I blowing wall warts with this but only after a few minutes?

[frequencycentral]

Weird. What is the spec of the power supplies you've burned out? How many milliamps? I recommend 500ma.

[Davidgwells]

The first was 12vdc 500ma, the second was 12vdc 600ma. 

[igerup]

An interesting thread, this one. I noticed that in the early stages of circuit developments (and your first build of this most interesting little amp) you connected the 0 ohm (ground wire) of the output transformer to ground. But you did not use an insulated output jack.  I know the B+ is lower than 100 volts and the amp most likely doesn't send 2.5 A to the speaker (like the averag 100 watter does). But output current will still be way higher than in the preamp, you don't want that in the preamp so use insulated output jacks. . The OT ground wire should only be connected to ground at the same place as the first filter cap.

For more info check http://www.aikenamps.com/ in the tech info section. There's some very good info on ground wireing there which also applies to submini tubes.

[criszou]

Hi, Rick
I'm sure you had record Murder One on computer soundcard. In this case, do you mic it, or use DI box, or direct to your soundcard?
What is Muder One's impedance for the load box?
I have collect  the kits one by one ^^. Do I have to concern about 110V or 220V? In here use 220v.

Thank's

[Cliff Schecht]

An interesting thread, this one. I noticed that in the early stages of circuit developments (and your first build of this most interesting little amp) you connected the 0 ohm (ground wire) of the output transformer to ground. But you did not use an insulated output jack.  I know the B+ is lower than 100 volts and the amp most likely doesn't send 2.5 A to the speaker (like the averag 100 watter does). But output current will still be way higher than in the preamp, you don't want that in the preamp so use insulated output jacks. . The OT ground wire should only be connected to ground at the same place as the first filter cap.

For more info check http://www.aikenamps.com/ in the tech info section. There's some very good info on ground wireing there which also applies to submini tubes.

I prefer to use the chassis as the ground for the speaker jacks, the speakers don't care and the chassis makes a great conductor for all of that return current. I keep this current out of the input sections by isolating the input stage (and usually subsequent stages) ground and only having this ground meet the others at the "-" side of the decoupling capacitor. This is physically the best place to run every ground to, but usually not the most convenient.

[merlinb]

I prefer to use the chassis as the ground for the speaker jacks, the speakers don't care and the chassis makes a great conductor for all of that return current.

Yikes! What awful advice. The chassis is not just one 'big fat wire'! You can get away with it, but don't make a habit of it...

[Cliff Schecht]

You think so? I don't rely on the chassis ground for everything mind you, actually I prefer to not use the chassis at all as a ground. The only place that I let the chassis ground touch the signal ground is where the speaker jack meets the chassis, but even this gets a direct connection to the B+ decoupling cap "-" terminal. Essentially the only place that chassis and signal ground touch is at the speaker jack. Keeps the noise WAY down in amps, at least the 60/120 Hz variety. Unless I'm doing something vintage inspiried, I typically decouple my amps like you show on your grounding page.

I guess the big point to make here is to know and control where your large return currents flow so you can keep them out of the input. The easiest way to do this is an isolated input jack, the output jack isn't anywhere near as picky because the SNR at that point is stupid large.

[merlinb]

The only place that I let the chassis ground touch the signal ground is where the speaker jack meets the chassis, but even this gets a direct connection to the B+ decoupling cap "-" terminal.

OK, at first it sounded to me like you were using the chassis to actually carry speaker current.

Incidentally, for lowest noise you want you only circuit-chassis connection at the input jack. (As you pointed out, the speaker is insensitive to noise, so you don't want to waste your connection there).

[Cliff Schecht]

The only place that I let the chassis ground touch the signal ground is where the speaker jack meets the chassis, but even this gets a direct connection to the B+ decoupling cap "-" terminal.

OK, at first it sounded to me like you were using the chassis to actually carry speaker current.

Incidentally, for lowest noise you want you only circuit-chassis connection at the input jack. (As you pointed out, the speaker is insensitive to noise, so you don't want to waste your connection there).

This is what Trainwreck's and some other amps do and it does work, but there are better methods out there. It's not as bad of ju-ju on an aluminum chassis but you're asking for trouble with steel because of the slightly higher resistivity and the fact that steel can couple EMI from transformers and such while aluminum doesn't.

[Davidgwells]

Anyone want to fact check my wall wart pick?  I have let my murder one sit, non working, for a few months and i'm ready to dive back in.  The link is for one I found on mouser, does this look like the right choice?

http://www.mouser.com/ProductDetail/Xicon/412-112054/?qs=sGAEpiMZZMtpkqKkT5w3ugVzZMH3McLBpaXAxv7xPGM%3d

[zambo]

It looks as if it is the right kind. I havent built a murder one though.