PYE Compressor/Limiter Thread *boards shipping* BOM up!

[desol]

I really like the way it sounds. It's really strange the way it does it's thing. Totally different than any of my other compressors. I get why some people freaked out over this thing (pye compressor). It can frame your mix (or acoustic guitar) in a really nice, classic reference...with just the right amount of punch. It kinda moves and sways in a unique way too....and has certain amounts of saturation I think (I could be wrong). You put it on your mix bus and it sounds like a classic record right away. It has a strong colour and good sustain.

 

[ruffrecords]

The way the compression works the low end on the drums is really cool on these units.

Reminds me of Pigmeat Markham song Here Come The Judge



Cheers

Ian

 

[desol]

Mine sounds so vintage'y it's crazy, and it's using the IC oscillator. I'm pretty sure I remember Abe saying that the oscillator type shouldn't really affect the sound of the unit.

 

[skal1]

ok the threads still alive ,


I have a few faults

the release is not working but it is compressing , and a 100hz bump , can anyone help me

cheers

denis

 

[skal1]

ok this is doing my head in why is my noise floor looking like this

 

[skal1]

bump bump. any body help

best

denis

 

[Ilya]

Looks like transformer induced hum. Try powering the unit from and external DC source (for example a bench supply) and compare the noise floor.

If it’s 100Hz (not 50Hz), then it may be a ground loop or other pollution with the rectifier noise.

 

[petermontg]

How's the required voltage and current looking? Too little?

 

[skal1]

Hi.

i have nt got external DC source, and the transformer out the case i am running to power supplies modules https://i.ebayimg.com/images/g/skgAAOSw5gVg9qxq/s-l1600.jpg ,i have never had a noise. problem with this kit

could it be too much current draw form a relay board? or is it the toroidal main transformer it is a two outputs 24ac i joined two legs to get the ground
pin the the power supply pcb

Best

Dennis

 

[skal1]

petermontg​

everything checks out i am running them at 16v right now 0.067Ma 2 pye boards. bypass board 0.111Ma

best

dennis

 

[skal1]

ok so i striped back the power supply and only used 1 to power the two pye boards and relay board and now i have nt got all that noise on the output..

Why would two psu in parallel cause the noise ?

i still have little hump at 50hz but thats ok because the noise floor is not rising when increasing gain

 

[Ilya]

Why did you have 2 PSUs in parallel? How did you wire them?

 

[skal1]

i took the ac secondary supply from transformer and wired them to both of the ac inputs on the external pcbs.

psu1 powered the pye pcb and pcb 2 powered the lights and bypass pcb

 

[Ilya]

Depending on how and where exactly you connect power feeds and grounds you may very well create a ground loop that shows up as a set of 50Hz harmonics (as in one of your posts). Without seeing the unit and tracing all connections it's very hard to say what is wrong.
General guidelines are:

• Connect grounds at one point. This should be a "clean" ground - that is, an output of a regulator, not a smoothing cap after a bridge.
• If you have several regs, connect them at their clean points.
• Don't feed DC power in parallel without balancing resistors.
• Connect your clean ground to the safety earth (enclosure) at a single point.

Maybe others will add to this list something else. My experience shows that these points are very important and help eliminate ground pollution in almost all cases.

 

[skal1]

cheers

Ilya​

 

[desol]

Here's another clip of my PI-3141 on a drum loop. It punches things up nicely. I have the input up about 1/2 of the way. There's more to go if you want. Cheers!

 

[maxwall]

Original schematic L1 & L2 shows 2 x 52.5mh in series, 105mh. Looks to be ferrite pot core types. The AC PCB only uses one ~ 50mh, instead of two at 105mh combined.








Any comments on these Wilco shielded inductors for L1 & L2 ~50mh and L3 (R38/100r) 12uh. Notes on side indicate not suitable for DC apps, due to low saturation.



Having trouble finding caps locally with correct lead spacing to fit the PCB. specifically the 150uf / 50v .

A option for less costly part MMBT3640 (sot 23, 3 leg) footprint to allow PN3640 pad flexibility , since TO-92 PN3640 is obsolete and pricey part. Possible Alternative to PN3640 (TR11) : MPS3640, 2N4209 (15v Ic 50ma ? ) , 2N5771 (15v, Ic 200ma ) , MPS5771 ( 15v Ic 50ma ?) Not sure what the min Ic current req should be , PN3640 is 200ma, and if 12v vs 15v Vce makes any real difference. If not maybe a 2N3906 or 2N4403. The PN3640 is a high speed switch , not sure if a general purpose PNP with similar Vce and Ic would work instead.



Would be nice to have a toroid or pot core inductor specified for the osc board as a alternative to the TLC555 chip to keep the design as close to original. Ex: wire # , core part number description (ferrite or permalloy) and wire turns starting point are unknown leaving TLC555 chip the only option , the cost on the chip is reasonable though. Should'nt be to hard to experiment with toroid cores and windings to get close. The one pictured ( T3 80uh ) is a small one for the oscillator board. Not sure if 80uh is accurate though. Time to gather some cores and wire for some experiments. It looks as if its a easy wind at around 3-4 turns.

 

[Ilya]

Any comments on these Wilco shielded inductors for L1 & L2 ~50mh and L3 (R38/100r) 12uh. Notes on side indicate not suitable for DC apps, due to low saturation.

These are a part of LPF filter. Shielded is good. If you look at the input and output of the filter, you'll se coupling caps. This means that the filter doesn't "see" any DC. You should be fine with these inductors.

As for transistor, look for switches that are capable of 400-500 kHz switching frequency minimum. This is a finicky part and you might need to check several different transistors to find a suitable one. Look for the DC offset at the output when engaging compression. This may be a sign that you need another transistor.

Can't help with the oscillator tx. Experiment.

 

[maxwall]

These are a part of LPF filter. Shielded is good. If you look at the input and output of the filter, you'll se coupling caps. This means that the filter doesn't "see" any DC. You should be fine with these inductors.

As for transistor, look for switches that are capable of 400-500 kHz switching frequency minimum. This is a finicky part and you might need to check several different transistors to find a suitable one. Look for the DC offset at the output when engaging compression. This may be a sign that you need another transistor.

Can't help with the oscillator tx. Experiment.

Thanks feedback , don't understand original two inductors L1 ~50mh and L2 ~50mh , why not L1+L2 combined for one inductor at ~100mh ?

Also , very painful to find 150uf 50v ( five required ) capacitor that will fit A/C Pi-3142 circuit board. PC board was overlooked in this area in the design phase. Lead spacing 3.5mm or less does not match common 5mm lead spacing needed to accommodate standard 150uf/50v capacitor. Lead spacing of 5mm minimum. Very very few miniature sized capacitor choices available with 3.5mm lead spacing at 150uf @ 50volts. Body of capacitor must be around 8x12mm to fit. Most are 10x12mm at 5mm leads spacing. Can I use a 120uf / 35v ( size 8x12 mm) or 180uf /50v (size 8x20mm) instead , since lead spacing is closer to 3.5mm. Any performance consequences with higher or lower capacitance ? It appears the original caps were 125uf / 25v

Anyone have experience with China JCCon ( black and gold ) electros for low esr audio signal caps ?

Lastly , regarding PN3640 , was thinking of using a surface mount adapter board for part TR11, MMBT3640, since the surface mount part is 1/10 the cost of thru hole part. Will the adapter board be a point for unwanted freq oscillations due to added distance from main circuit board with surface mount adapter board ?
The adapter board converts surface mount part to T0-92 three leg.

Suppose my curiousness means bench testing for myself

 

[ruffrecords]

Thanks feedback , don't understand original two inductors L1 ~50mh and L2 ~50mh , why not L1+L2 combined for one inductor at ~100mh ?

Because the filter consists of a pair of cascaded PI filters used to filter out the switching frequency and leave only the audio.

https://www.elprocus.com/pi-filter-circuit-working-and-its-applications/

Cheers

Ian