unlikely it needs heatsinks, or just really small ones.

I've noticed that the BOM for this, as well as for some of the KGSSHV Carbon stuff, uses large value tantalum caps (47uf/25V), for which there are also ceramic MLCC X7R alternatives available. Any reason for using the tantalums as it would seem that the ceramics would have advantages, price being one of them?

For example the tantalum 47uf 25V 581-TAP476K025CCS @$3.13 could be replaced with a Murata 81-RDEC71E476MWK1C3B  @$2.13 MLCC ceramic.

Just curious...

Edited September 28, 201510 yr by Pars

Thank you so much for this Kevin 

Might you have a link to the updated schematic?

I've noticed that the BOM for this, as well as for some of the KGSSHV Carbon stuff, uses large value tantalum caps (47uf/25V), for which there are also ceramic MLCC X7R alternatives available. Any reason for using the tantalums as it would seem that the ceramics would have advantages, price being one of them?

For example the tantalum 47uf 25V 581-TAP476K025CCS @$3.13 could be replaced with a Murata 81-RDEC71E476MWK1C3B  @$2.13 MLCC ceramic.

Just curious...

aren't the mlcc's microphonic?

They can be but tantalums also have a tendency to explode so I'd use the ceramics... 

only if you stick them in backwards.

(or if they are labeled backwards which I have seen)

updated schematic... you gotta be kidding. probably this weekend.

and then I am going to have to show how you calculate the trimmer resistors for other than 20v

I just recall the wet tantalums back in the '80s would blow'd up real good, bubba And start on fire, with silver shit coming out of them.

Generated a distrust in them, to say the least.

same question here,is there any reason to use tantalum caps at certain places like C8、C13 instead of electrolytic or ceramic caps ?

Any reason for using the tantalums as it would seem that the ceramics would have advantages

 

someone with a screen room and a 8.5 digit dvm needs to test 

I have access to both    

well, what are you waiting for.

Too busy, hey it's your fault Kevin, you keep coming up with these amp designs that are eating up my free time 

lousy excuse.

before I retired, I used to work with Agilent 3458A multimeters

Great for Dc voltages, but for AC you need a Wavetek/Datron 4920 or a Fluke 5790A/B.

 

schematic updated with parts for latest board

 

schematic updated with parts for latest board

 

Link the same as the previous?

 

EDIT: Also, a couple of questions:

1) The 1N4007 diodes are not shown in the schematic. Could I use 1N4001/4002 here (have them on hand)? I'll trace them out on the board to figure out where they are, unless they are in the updated schematic.

2) The 2K resistors (R13/R14) and the LED between the + and - output: what is their purpose? I noticed that Kevin had not put the LED in, and others completed boards as well were missing these. A dummy load for testing?

Edited October 11, 201510 yr by Pars

Link the same as the previous?

 

EDIT: Also, a couple of questions:

1) The 1N4007 diodes are not shown in the schematic. Could I use 1N4001/4002 here (have them on hand)? I'll trace them out on the board to figure out where they are, unless they are in the updated schematic.

2) The 2K resistors (R13/R14) and the LED between the + and - output: what is their purpose? I noticed that Kevin had not put the LED in, and others completed boards as well were missing these. A dummy load for testing?

1N4007 is for protection as shown below, 1N4002 would be better if you only have 4001 and 4002

 

LED between + - rails can be power indicator and simple load to help discharging caps.

Thanks. I had figured the 1N4007s were there to protect something, but hadn't ohmed or traced them out yet. And "DOH" on the LED/2K resistors

schematic updated with parts for latest board

 

Kevin had noted an updated schematic, but what I see in the schematic posted previously still shows mj15024/15025 TO-3 devices, no protect diodes, etc.

updated schematic with diodes and right output transistors.

(at least for the absolutely latest boards)

Thanks Kevin!

At first I did a WTF with the 3 bridges until I figured out why

Another GoldenReference LV Dual lives! This one is, obviously, built with the the board from the Carbon GB. 

Had it running for 30 minutes for testing and the regulated voltage was rock steady as kevin mentioned. I hand-matched the resistors that determine the output voltage so I am annoyed that I am getting +/-  14.99VDC. I blame it on the LT1021-10 reference. I got the 0.5% tolerance ones so the result is within range (why can it not error on the other side?). I'll decide later if to go through the trouble of desoldering the change the resistors to get to +/- 15VDC. 

The heatsink is more for show. I plan to use this one to supply the LV for the HV Carbon. I doubt it they'll get any warmer than room temperature.

Thanks Kevin for the great creation.

Edit: forgot to mention, I used 2.1V LEDs and smaller bridge rectifiers. Just happened to have these parts on hand and the under-sized rectifier's spec'ed output current works for my forseable applications.

 

 

Edited October 12, 201510 yr by mwl168

That is massively OCD close enough on output voltage

 

BTW, your meter seems to be missing a knob

Yes, the missing knob is a long story. Basically the shaft of the knob just snapped when I went to turn it on one day. Probably exceeded the maximum turnning velocity and torque combination specified by the manufacture. So please treat your meter gently. 

Took it apart but no replacement part available from Fluke. I complainted and they were nice enough to offer me a new one with special discount. 

The meter still functions fine, so I got a Torx 27 socket and use it as the knob. Works really well and has the added benefit of preventing unauthorized use of the meter.

Having worked in calibrating multimeters, among other things, your problem could be not enough resolution in the meter.

Solution: Use the voltage range closest to the voltage being measured without going under the voltage being measured, or get a multimeter with a greater number of digits and the accuracy to match.

To measure 15.00V with only 4 digit resolution gives only 2 decimal places.

 

If the voltage is exactly 15.00V, the the reading could be anything from 14.99V to 15.01V depending on the resolution and the tolerance of the meter.

With more digits and resolution you could get between 14.99999999V to 15.0000001V and still be 15.0000000V

Thanks Wink. 

I did not consider the toleance of the meter. I will check and see what range adjustment my meter offers.

Any recommendation of meters that are more accurate and has more resolution without breaking the bank?