another 555 smps layout

[Marcos - Munky]

I built a Marshall 18W amp from a really cheap kit several years ago, and the supplier put the wrong power transformer in, so instead of getting 270VAC I was getting about 500VAC from the secondary. The amp sounded ABSOLUTELY AMAZING up until the tubes melted. I never measured what B+ I was getting, but I'm sure it was absurdly high.

Ouch . The tubes surely lasted too long. Was the retification done by diodes? If yes, then it could be they messed up with the rectification part of the circuit. A 500VAC (250V + 250V) transformer with center tap connected to ground and two diodes for rectification works the same as a 250VAC transformer with no center tap and a four diode bridge rectification (if we don't take current losses into account). But they could messed up with this part of the circuit and used a transformer with no center tap and for diode rectification, but with 500VAC as the secondary voltage instead of 250VAC. Well, 250V is a low voltage for a 18W, but still makes way more sense than 500V .

Just by curiosity, back on 70's a brazilian brand called Giannini made some amps called Duovox. They made models with a pair of 6L6GC, a quartet of 6L6GC and even a quartet of KT88. While they sounds very good and are very powerful, they're known for melting the 6L6. The reason? The nowadays 6L6GC can handle up to 450V-500V, but those amps sends 740V to the 6L6GC plates! The thing is they were made back then when the RCA 6L6GC were very easy to find, and this exact tube can handle this higher voltage. A good nowadays tube still can handle this voltage for some time, but a cheap 6L6 won't last for a few minutes . So any owner of this amp needs to have the power transformer replaced if they intend to keep using this amp, or search on ebay for a expensive set of RCA tubes.

[Elijah-Baley]

So I can set the max voltage in my tube preamp from the SMPS board, and the voltage tube stages thorugh the... load resistors, I guess? Is it called like that?

I got that the voltage in the tube preamp depends from the power supply and the tube preamp schematic and it acts on the headroom.

[Marcos - Munky]

So I can set the max voltage in my tube preamp from the SMPS board, and the voltage tube stages thorugh the... load resistors, I guess? Is it called like that?

Yep. But you can call them plate resistors, since they're connected to the tube plate. they're basically the collector resistors, but for tubes.

You can even do a hybrid design where the tube gets powered by let's say 200V, and a resistor drops the voltage to let's say 20V to power a fet.

[Elijah-Baley]

Drop to 20V for a JFET buffer or a booster, maybe.

[vigilante397]

Drop to 20V for a JFET buffer or a booster, maybe.

I started doing that with my Kingsley Harlot clone, uses  JFET to push the tube. He drops to about 27V for the FET if I remember right.

[Twhjelmgren28]

This is a fantastic thread and great timing for me!  I'm working on a Harlot clone and fell down the SMPS rabbit hole a bit.  I meant to post here sooner but had to gather my thoughts first.

In reading this thread, in addition to this one:  https://www.diystompboxes.com/smfforum/index.php?topic=124809.0 - in which Marcos - Munky has tons of great posts, I have a few follow-up questions.

I was hoping to make a sort of list to figure out what parts play the biggest role in these circuits.  I've made a list of the key parts with summary of key points made by all you guys along with my silly questions:

1. Mosfet - with low Rds runs cooler which gives higher current - but can't handle as much voltage; mosfets with higher voltage ratings also have higher Rds meaning they run hotter (so require a heat sink) and give less current.

2. An ultra-fast diode is required - The faster the reverse recovery time, the better - current rating is a factor

3. Inductor - Is the main concern the current / voltage rating?  Does changing the inductance value alter the circuit much?

4.  Layout is hugely important for minimizing noise and allowing space for a heat sink - as Marcos notes in the OP

5.  Does swapping transistors from BC547 have an effect?  Or again, is this mainly a current / voltage rating?

6.  Capacitors - Marcos mentioned in the other thread that a lower ESR achieves higher current.  Any other thoughts with this?

7.  Resistors - I know a lot of these play a significant role in setting the timing (is that the right word here?) of the switching.  Do any certain resistors play key factors in affecting voltage / current?

8.  Do we need any other filtering following the main SMPS circuit?  The Harlot clone appears to have some filtering between the power supply and JFET but I'm not seeing any filtering between the power supply and tube plates.  I thought that was interesting.


I know this is a heckuva list and I'll be doing some experimenting on my end, after finishing up the Harlot clone, but I thought I'd ask you guys first.  I've done a ton of looking around the web related to SMPS and while a lot of sites use the same circuit, there isn't a ton of explanation on how it works.  I believe the original article that explains it is the best I have found, however, I'm interested to know if you guys have any additional insight. 

[Marcos - Munky]

Glad to know it's useful for you. So let me try to reply your key points, based on what you want to build.

1- that's correct. But the Harlot uses just a single dual triode tube, so you just need a bit of current. I'd just use a IRF740. Btw, a higher Rds may require a heat sink, or may not. And you can just mount it on the enclosure (be sure to insulate it) and call iot a day.

2- just go with a UF4007. It's more than enough for guitar pedals.

3- current is a concern, even when you're powering a low current circuit. I saw some people using inductors rated for less than 2A, but it's better to be safer and use a 3A one. As for inductance, it can mess with the switching frequency, so I'd keep it as is.

4- from what I remember, parts placement are important. Mostly, the mosfet, inductor, diode and IC must be as close as possible. But you probably won't need a heat sink for a single 12AX7.

5- no difference at all. The original schematic calls for a BC547, I just used the 2N5088 for the layout because I have lots of them and just a few BCxxx. In fact, the one from the photo used a BC547. Just remember to rotate them in 180º when swapping one for the other.

6- low ESR caps indeed does a significant improvement for higher current circuits, like small amps. But I can't find them down here, so I just use what I have on hand. Your circuit requires a low current, so you'll be fine with regular caps. Worst case, the mosfet will heat a bit more to keep up with the current, but I woudn't bet on that for the Harlot.

7- current doesn't really depends on the resistors. R5 and R6 (on my layout) are related to max and min voltage. You can play with them a bit, but they're ok as is.

8- usually the preamps themself already have the required filtering on their schematics/circuits, and the smps have a bit of filtering. You can indeed add more filtering, adding a big-ish electrolityc cap (47uF to let's say 220uF) in parallel with a 100nF cap, both connected between high voltage and ground. I'd say to test it as is first.

Btw, the Harlot is on my "maybe I'll give it a try" list for some time. It's a very nice sounding overdrive.

[vigilante397]

+1 to everything Marcos said. My comments on a couple points though:

3. I build in bulk so I buy in bulk, and to save money I started using 1.21A inductors in the SMPS and haven't had any issues, but obviously a higher current rating will never hurt and is always a good idea.

6. It's not so much that low ESR achieves higher current output from the SMPS, it's just that the more current you pull the more ESR starts to become important. If you're using a couple 12AX7 you're talking about single digit mA, so just about any cap rated for the voltage should be fine.

8. What schematic are you looking at for your Harlot? The only one I'm aware of (the one I traced from a Harlot v2 last year) doesn't show the SMPS at all. If I remember right (I'd have to look through my pictures) there was a ferrite bead between the SMPS and the circuit, but then again it may have been between the power input and the SMPS, I don't remember. I don't use any additional filtering in my Harlot builds.

[Twhjelmgren28]

What schematic are you looking at for your Harlot? The only one I'm aware of (the one I traced from a Harlot v2 last year) doesn't show the SMPS at all. If I remember right (I'd have to look through my pictures) there was a ferrite bead between the SMPS and the circuit, but then again it may have been between the power input and the SMPS, I don't remember. I don't use any additional filtering in my Harlot builds.

I'm looking at the schematic on freestomboxes.org.  The same one is also on PedalPCB.  Do you have another schematic?  The filter I'm referring to is not physically between the power supply and the FET but electrically - there is a 330K going into a 4u7 that's attached to ground.  The 330k also appears to be part of the voltage divider for the FET.  I could absolutely be mistaken; that's just what I thought was going on...

[vigilante397]

I'm guessing it looked vaguely like this? The 330k is part of a voltage divider to drop the HV line down to 27.4V (what I measured from the original unit, your voltage may vary) for the JFET. While it does form a low-pass filter with the 4u7 cap that isn't really it's primary function.



Plenty of stuff to ignore in this as it's specific to my builds, like pin numbers on the tubes, SMD part numbers, bridge rectifier on the input, etc.

[Twhjelmgren28]

I'm guessing it looked vaguely like this? The 330k is part of a voltage divider to drop the HV line down to 27.4V (what I measured from the original unit, your voltage may vary) for the JFET. While it does form a low-pass filter with the 4u7 cap that isn't really it's primary function.



Plenty of stuff to ignore in this as it's specific to my builds, like pin numbers on the tubes, SMD part numbers, bridge rectifier on the input, etc.

Ahh - thanks for posting that!  Actually I think it is your schematic.  Looks like Megatrev posted it on Pedal PCB and you piped in further down the thread.  I did see that the 330k and 68k form the voltage divider but I assumed the only reason he added that 4u7 was for filtering...

I actually do have questions about what specific parts do in this circuit.  There are a couple strange things I haven't seen before.  Is it okay if I PM you so I don't bring this thread too far down the rabbit hole on Harlot clone discussion?

[vigilante397]

Ahh - thanks for posting that!  Actually I think it is your schematic.  Looks like Megatrev posted it on Pedal PCB and you piped in further down the thread.  I did see that the 330k and 68k form the voltage divider but I assumed the only reason he added that 4u7 was for filtering...

I actually do have questions about what specific parts do in this circuit.  There are a couple strange things I haven't seen before.  Is it okay if I PM you so I don't bring this thread too far down the rabbit hole on Harlot clone discussion?

Yup, you can PM me with any questions.

Megatrav was the one that bought the Harlot and sent it to me to trace, so I thought he deserved a copy of the schematic