Finished capacitance multiplier

Status
Not open for further replies.
I Like It

One down...One to go.😀
 

Attachments

  • SAM_2010.JPG
    SAM_2010.JPG
    657.7 KB · Views: 452
Thanks Rick. Transistor ZVN2106A is backorder from Mouser. ..
I should have some spares from when I was making the kits for a few people who don't have good (cheap) access to the usual distributors. Last time I mailed a few parts in the US, postage was $1.17. Not sure how many are left, but I can also add a few to my next Newark order if I don't have enough. I have an order going to Newark/Farnell about Wed, I think.
 
Any reason why mica washers are specified for the MOSFETS? I haven't bothered insulating the transistors since each one has its own heatsink. Anyone see a problem with this?
I usually specify insulators for heatsinks. If someone wants to leave them off, at least they have been warned, so to speak... 😀

Actually, any kind of insulator would do the job, I did not intend to specify mica. Main benefit is electrical isolation from the heatsink in case of shorts, for example when someone is setting or checking the pass transistor voltage drop.

For best results, you should at least check this with the amp as a load, and make sure the positive and negative rails are the same. This normally means when installed in the chassis, with limited space, so a short is a definite possibility.

You may have also noticed there is a fuse cover in the BOM, although most of my pics do not show one installed... 😉
 
Only on the faston tabs. I had one pull through previously when pulling off a particularly tight connectors.
A cold solder joint can cause that, but one of the main reasons is holes that are not plated through. Without hole plating, solder will not fill the hole, and the pads alone cannot possibly hold that much force by themselves.

The solder in the hole makes the connection between solder on top and on the bottom. Just like the center part of an I-beam in building construction. Take that away, and there is not much strength left in the joint.
 
OK, So What am i doing Wrong Here. I have a 120V Pri traffo With 2 30VAC Secondaries.
I only have 1 working VSSA module at present & 1 Cap Multi. So i hook up one Sec (blue/Grey) to AC1 & AC4. I also have 1 green 'Screen' Wire, so i put it on AC2. With NO load i am getting only 21VDC at Fuse. I can adj the output to be within .5V of this reading.
The AC across AC1&2 is still 32.5VAC. So why is DC voltage so Low?
 
OK, So What am i doing Wrong Here. I have a 120V Pri traffo With 2 30VAC Secondaries.
I only have 1 working VSSA module at present & 1 Cap Multi. So i hook up one Sec (blue/Grey) to AC1 & AC4. I also have 1 green 'Screen' Wire, so i put it on AC2. With NO load i am getting only 21VDC at Fuse. I can adj the output to be within .5V of this reading.
The AC across AC1&2 is still 32.5VAC. So why is DC voltage so Low?
The way you have it hooked up, you are splitting the secondary AC voltage. Better if you use both secondary windings, one per rail. In that case, you should see around 30 * 1.414 ~ 42V on each rail (plus about 5% at no load.)

If you have 4 wires from the secondary side (they should be color-coded), you should always connect one sec. to AC1-AC2, and the other secondary to AC3-AC4. If the transformer windings are marked with a black dot on one end, then this wire should be connected to AC1 and AC3, respecively, for the two secondary windings.

I do this with the fuses out, and then I check the DC voltage at the fuse sockets, and make sure I have about 1.4-1.5 times the nominal secondary voltage, both positive and negative rail.
 
Status
Not open for further replies.