the Poor Man 660 support thread

[khstudio]

[quote author="khstudio"]HEATSINKS - MOUSER

Q4, Q5
Mouser Part #: 532-529902B25G
Manufacturer Part #: 529902B02500G
Manufacturer: Aavid Thermalloy
Description: Heatsink TO-220, VERT 3.5 TR
Case Type: TO-220
Dimensions: 41.91 mm L x 25.4 mm W x 50.8 mm H

IC1, IC2
Mouser Part #: 532-531102B25G
Manufacturer Part #: 531102B02500G
Manufacturer: Aavid Thermalloy
Description: Heatsink TO-220, VERT 10.4 TR
Case Type: TO-220
Dimensions: 12.7 mm L x 35.052 mm W x 38.1 mm H[/quote]

 

[mitsos]

OK, yous guys... I gotta problem. one more trafo question for a MONO PM. I am trying to do this the G9/G7 way, as Niels suggested in another thread, but before we knew what the 18V was for. Ideally I would like to have relays but it is not a requirement at this point. My problem is I don't know if I can get all the necessary voltages, mainly the -17... So I took some pictures of what I think may work :?

EDIT: It just occured to me that I could use two toroids with equal V secondaries :roll:

My idea is to use Antek toroids, the first idea is 1209(2x9V, 100VA) and 0506(2x6V 50VA) backwards and the second idea is the same but using two 0506 trafos.

In the first, I have 9V 2.78A which would be 12+ rectified and which has to be brought down to 6.3 without regulation. Possible?

Also, is 9V 3A going into the secondaries wired in series of the second going to blow up that trafo? If not, what voltage can I expect on the primary side? The secondaries are 6V 3A each. And can I tap that like the G7 PSU to get -18V (with a doubler)? I suppose I can't get the +18 out of these trafos, can I? Click the pic to see what I mean. The voltage doubler would be on the 9V that feeds the 2nd trafo.




Fearing the above might be dangerous, I thought maybe I could use the 0506 connected to mains to give me 2x6V 3A, and wire one secondary to another 2x6V in parallel (50VA maybe not necessary), thinking that this should get me about 250 on the other side? I thought to use the same voltage multiplier idea(this time to triple V to -18?)

Am I on the right track? Is there a better way without buying an expensive custom trafo? Is there ANY way to get both + 18 and -18 from this type of thing? Could I feed two voltage multipliers from the same secondary?

Thanks!

 

[cannikin]

[quote author="emrr"]So Cannikin can confirm Mouser heatsink #s!!

My boards showed up 8/9. I started stuffing this afternoon. Still more parts to order.[/quote]

khstudio posted the right ones, for IC1, IC2, and Q1 on the PM board I used 532-529802B02500G I had some already in stock. they are the same ones, only KH's are a little taller.

 

[emrr]

This attenuator calc, found in the software META, is as detailed and easy to use as any I found in a brief search. You still have to come up with nearest values yourself.

Resistor string value calculator for constructing stepped attenuators:
http://homepages.tcp.co.uk/~nroberts/atten.html

This one works fine in Explorer for me, but not in Firefox or Safari, and the final resistor value is not shown, but assumed from the string value. Less manual labor with data entry here, still have to do nearest values yourself.

http://www.quadesl.com/attenuator.html

 

[analag]

Sorry guys, I try to keep up with the thread but work has been demanding.
Remember IRF820 for mono and IRF840 for stereo.

analag

 

[khstudio]

Deleted

 

[Rob Flinn]

Anyone got a uk/euro source for the wima 10u ????

 

[flintan]

Conrad no. 450535. It's in the parts list as well. :wink:

 

[khstudio]

Deleted

 

[khstudio]

[quote author="[silent:arts]"]the rotary switches

the first rotary switch is the "Input" or "Gain" switch.
dual deck is necessary, shorting preferred, and the more positions the better. over all resistance is 15K. matching the resistors for each step on both decks is a very good idea.
how to calculate the resistors ?

if you want it linear this is easy: 15000 / (number of positions -1) = single resistor in R
(you need (number of positions - 1) resistors of the same value)

you would better want it exponential (or log in potentiometer language).
it took some time to figure this out, but with some help I was able to do an Excell sheet to calculate the resistors for emulating a real log potentiometer.
fill in the desired steps you want: Log Pot to Switch Excell Sheet

you may want it in dB steps (like me). at the end it is a series attenuator.
my prototype has 18dB of gain, thus I will do one dB steps on an Elma, stopped to 23 steps, going from -4dB to +18dB.


the second rotary switch is the "Threshold" or "Compression" switch.
same as the first one, but the dB step attenuator doesn't make sense here. Lin or Log, your choice.[/quote]

I don't mean to beat this to death but I need to understand the LOG switch better & this chart. + I don't have the schematic yet but i am ordering parts & want to make sure. The work involved in building these is too much to not ask first.

In the excel sheet it shows position 1 with "0.00" resistance. :?
Is this right? I'd think that ZERO Resistance would let the FULL signal thru... shouldn't this be the LAST position :?:
If so... is the TAPER backwards on the Excel Chart???

I'm still wondering (since I haven't actually used the PM670 YET) which taper would suit the "Threshold" best?
I seem to lean towards the LOG for more refined adjustment on the Lower Ratio... not sure here.

 

[khstudio]

From the Excel Sheet - 12 STEPS

http://www.twin-x.com/groupdiy/albums/userpics/12-Stepped-Gain-LOG-1.jpg

From: Doug... it Did work for me in Firefox 3
[quote author="emrr"]
This one works fine in Explorer for me, but not in Firefox or Safari, and the final resistor value is not shown, but assumed from the string value. Less manual labor with data entry here, still have to do nearest values yourself.

http://www.quadesl.com/attenuator.html[/quote]

NOTICE- The Resistors are "Backwards" compared to the Excel Version... BOTH are suppossed to be LOG Taper. :!: :?:

 

[emrr]

Threshold is an interesting problem; on one hand log makes sense, on the other realize you are controlling signal levels that have probably already been compressed, so the relationship to 'pure audio' levels is already somewhat out the window.  Especially with heavy limiting, log steps might feel large.  But then, Analag says he's using 12 steps in both positions with no troubles, so....

Realize that if you do the threshold atten exactly like the [silent:arts] gain, then there is no 'compression off' position in the amp.  There are various solutions for that.

Another thought:  matched 1M resistors from attenuator sweepers to center of the two 15K arms would give safety against attenuator failure and open grid condition.  

 

[emrr]

Run this one:

Resistor string value calculator for constructing stepped attenuators:
http://homepages.tcp.co.uk/~nroberts/atten.html

Gives more descriptives in the results to help interpret outcome.

 

[[silent:arts]]

[quote author="khstudio"]I don't mean to beat this to death but I need to understand the LOG switch better & this chart. + I don't have the schematic yet but i am ordering parts & want to make sure. The work involved in building these is too much to not ask first.[/quote]
look at the block diagram, you do not really need the schematic to see how the attenuators work.
[quote author="khstudio"]I'm still wondering (since I haven't actually used the PM670 YET) which taper would suit the "Threshold" best?
I seem to lean towards the LOG for more refined adjustment on the Lower Ratio... not sure here.[/quote]
nobody can tell you which you will like best.
but you can start using potentiometers, decide which law fits your needs and build the switches afterwards to your needs.
[quote author="emrr"]Realize that if you do the threshold atten exactly like the [silent:arts] gain, then there is no 'compression off' position in the amp.[/quote]
the lowest switch position is connected to GND, there will be no signal in the sidechain amp, which I would say is "compression off".

 

[khstudio]

look at the block diagram, you do not really need the schematic to see how the attenuators work.

:thumb:

OK - Back to my point about the INPUT GAIN:
Why is the first position 0.00 Ohms?
Shouldn't this be the LAST Position?

but you can start using potentiometers, decide which law fits your needs and build the switches afterwards to your needs.

YES, good point... I've been looking for Dual 15k's to do this first but no luck yet.

Anyone have a source?

 

[[silent:arts]]

[quote author="khstudio"]OK - Back to my point about the INPUT GAIN:
Why is the first position 0.00 Ohms?
Shouldn't this be the LAST Position?[/quote]
this depends from which side you look at the switch :green:
look closer:
your pin1 of a pot is your first position on the switch.
your pin2 of a pot is the wiper
your pin3 of a pot is the last step of the switch.
now we start measuring from pin1 to pin2, what do you get at the first switch position ?
[quote author="khstudio"]I've been looking for Dual 15k's to do this first but no luck yet.[/quote]
you won't find one, for testing the law just take 10K (stereo)

 

[khstudio]

OOOOOOOOOO Yea... you guys are gonna love me (or kill me) now. :razz:

Are 1/4 watt Resistors large enough for the Threshold?

Just asking because they're after the the tubes.

 

[[silent:arts]]

I used 1/4W resistors with no problems yet.

 

[khstudio]

this depends from which side you look at the switch

Ahhhhhhhhh I see now. :green: :thumb:

 

[Rob Flinn]

[quote author="flintan"]Conrad no. 450535. It's in the parts list as well. :wink:[/quote]

Thanks, I didn't realise that Conrad was a European Supplier, I never heard of them before. They also don't seem to have an option to look at the site in English unless I'm missing something. Even there international page comes up in German !