Lm4702

[dfdye]

Alright, time to get the ball rolling on making some PCB's. Since I am still REALLY new at this, I figured I would run a design by you guys to see if I am anywhere near the right way of thinking about this.

The resistors are all vertically mounted to allow a variety of different sizes/brands to be used. The thinking was that using 1/2W resistors would be easier with this configuration without taking up a ton of real estate on the board. I know the silk screen layer overlaps in several places, but I am planning on etching these at home and not applying a silk screen. (however, is there a way to move labels in Eagle without moving the whole component?)

After the discussion over using bipolar electrolytic input caps and bypassing them with film caps, I put in space for using 1uf film caps paralleled with 33uf electrolytics. I noticed on Russ’s boards (which, with apologies to Russ for my bad design , was the starting point for this board) that there are larger electrolytics on the power inputs that I am not 100% sure the purpose of (are these additional reserve caps, or are you putting in power directly from the transformers here and using these as smoothing caps?) The parts I have tentatively picked out for the comp caps (CC1&2) are ceramic, though if there is some reason I shouldn’t use ceramic in a circuit, I would love to hear some alternate suggestions!

Thanks!!!

 

[dfdye]

Woops! Connected the CC's wrong! Major screw up.

Anyway, I think this board will at least be connected right.

 

[Nordic]

let us know when your prototype runs, my 2 chips are waiting patiently...... Many thanks to the member who donated them.

 

[dfdye]

I will definitely let you know when the prototype is up and working (hopefully some time within the next week if all of the parts get here, but that may be a little optomistic!), but I am having a hell of a time getting some of the more fine traces to transfer to the blank PCB's. So far I have used Epson Photo paper, Laser overhead and inkjet overhead material as toner transfer material, and I think that I just need a little more practice to get the fine traces around the chip to come out clean. For now, it is all sharpie!

I'll be etching the first board this afternoon.

 

[Nordic]

I found a way of making my magic marker traces survive the acid... I go over it using some enamel I used for a model plane...

 

[dfdye]

Thanks for the tip. The etching didn't really go that well, so I am definitely looking for a way of beefing up the traces that don't survive the initial transfer.

David

 

[Ipanema]

LM4702 + DS Blameless output

Guys,

Can this configuration work? I just plug-in Douglas Self EF output stage. The board is seperated into one LM4702 driver and two output stage. Pls advise.

Regards.

 

[Ipanema]

Output board

 

[dfdye]

Interesting suggestion. I am not sure if anybody has a working amp up and running yet (though I bet Russ has by now), but as soon as I get my transistors in, I will start playing around.

In the mean time, I have ruined several blank PCB's trying to get good home brewed boards.

Olimex here I come!!! Oh, and I fixed/tweaked a few more items on the boards. Once I'm happy with the final versions, I'll post in case somebody wants one.

David

 

[Ipanema]

Here's the layout :

Driver board

 

[Ipanema]

Output board layout :


I've made the board. When tested, I don't know why the bias current keep on increasing from 50mA to 300mA until I switch OFF the power supply. Anything wrong with the design? Pls help.

Thanks.

 

[dfdye]

Dumb question, but did you make the LM4702 package in Eagle or did you grab it from some where? It looks quite nice (a lot better than the one I made! ) If you did grab them, could you point me toward where you got them?

Thanks!

 

[Ipanema]

hi dfdye,

I made the layout myself but anyway it isn't working.

Regards.

 

[jackinnj]

Is the BD139 on the heat sink with the output transistors -- and how have you cooled (heatsinked) the LM4702?

 

[jackinnj]

I just haven't been able to find any film cap in 10uf that size anywhere, not just in US.

wrt 10uF input caps -- MKS suggested a few week back --

the usual suspects for WIMA's are Mouser and Future -- neither have this bird (MKS 10uF) in stock -- ASC's are pretty big -- but very good quality -- this cap doesn't have to be placed on the input board -- Audio Research places it on a pair of standoffs adjacent to the RCA input jack on their D120 amplifier.

You can occasionally find high cap ASC's and General Electric polycarbonates on EBay. Expect to pay over $10 a piece if you have to go through Allied or Newark.

 

[Ipanema]

Hi Jackinnj,

LM4702, BD139, MJE15034/MJE15035, TIP3055/2955 (currently used for testing) is mounted on the same heatsink.

The power supply is +/-34V. The heatsink is cool when bias is set to 30mA and it just keep on increasing until I switch it off at around 300mA. FYI, i've build a Douglas Self Blamesless amp before without having this problem. Pls help.

Regards.

 

[jackinnj]

ok

first, the LM4702 is the input and VAS stage and it probably shouldn't go on the same heat sink as the Output Stage -- i don't think that this is a major difference. the LM4702 is going to dissipate about 25ma times the absolute value of the rail voltage so it doesn't require a large heat sink --

the LM4702 is capable of sourcing/sinking some tens of milliamps -- so I trust that your output transistors are set up in Darlington configuration -- the beta of the 2955 and 3055's is comparatively low -- the MJ11032/33 has a beta of around 1,000.

i wonder a bit whether the resistor values specified by the Nat Semi pdf for the VBE multiplier are correct -- I would think you should have a trimmer so that you could adjust the current...i will try this out in the lab tomorrow.

the knowledge base on these driver ships is just being developed -- I note that Digikey shipped me some 4702A's which are supposedly still in the developmental stage --

 

[dfdye]

Now that I am getting close to putting together the output stage of the amp, I was wondering how Russ and others are using the thermally compensated transistors. I know that the SAP16's were suggested, but I wasn't really sure how these would be wired up. If Russ or somebody else could post a simple schematic of how these would be used, I would grately appreciate it! (Still learning!)

I am still planning on using regular transistors for the first itteration, but I would love to have some more options for future tries.

Thanks in advance!

 

[Russ White]

If Russ or somebody else could post a simple schematic of how these would be used, I would grately appreciate it! (Still learning!)

I am still planning on using regular transistors for the first itteration, but I would love to have some more options for future tries.

Thanks in advance!

Hi dfdye,

I have had my 4702s for quite some time, and have been working on a couple of possible layouts, but I have been quite busy over in the Pass Labs forum with a Zen V9 and my Twisted XBOSOZ, so I really have put this project on the shelf for now, I expect to get back to it shortly though. I think using the SAP16 will be one of the simplest most straight forward ways to get one of these going. It is a darlington so driver current demands are low, and it is thermally protected and has a minimum of external components. I will post more on the subject as time permits.

Cheers!
Russ

 

[hayenc]

If you look through the data sheets for the SAP 16's, it is not clear much more than the 200 ohm pot to set current is needed. I agree with Russ, it could be a very simple circuit for the 4702's.