8PAKAMP & 12PAKAMP KIT BUILD THREAD

[Joe Malone]

I ended up using 476ohms (2x238 in series) but im using un balanced so might have to drop that to maybe 238 to get back up to around 27-28db gain. Will have to measure it once I am done.

Much the same as my 5 channel version but with a 25-0-25 500va as no bridging going on here. Using 2 x 8 way Speakons for the speakers to connect to my Linkwitz Orion's

Excellent. I really like the layout of your amps. You will find that you will get better performance (lower THD hum and noise etc) if we take the 0v -speaker wire from the provided pad holes on each amp on the PCB ground plane. This is due to the star ground being before the capacitor bank which is on the 8PAKAMP PCB. So the current in the 0v power wire going to the PCB modulates. So the smoothest best place to take the 0v is on the PCB where the cap bank 0v ground plane is. This would make your star ground a lot smaller as well.

About to use a lot of these 8PAKAMP'S in a new 1000 seater theater for a full dipole speaker system that is huge.

[wizardofoz]

Well the bad news is on power up I have a 25v positive to 25v negative voltage swing on all outputs on every speaker output to ground. Some of them appear to swing back and forth with no inputs connected or loads...

Edit...actually they all swing...but all play absolutely fine when they have an audio source connected.

Update .... Well if I connect a source on the input then the output works for that channel and is fine so I assume its going into massive oscillation on the unterminated inputs. Never seen this on my other 8 channel build. Only difference is I used wima 0.1 caps on the bypass caps and 1k resistors for the gain setting. And this build is 25-0-25 where as the other one is 18-0-18 AC as I am not bridging any channels, this build is unbalanced inputs only and 478ohms for the gain. Pin 3 is grounded at the inputs for all channels.

Joe is there a need to have the trim pots installed and fully open for solving the oscillation issue or just a restistor of some value across the inputs or on the boards...if so where would i put the resistor and what value?

Any thoughts anyone?

[wizardofoz]

No thoughts or insights anyone before I gut this install and attack it another way?

I'm thinking to drop the 1k resistors down to like 238 as with unbalanced i need a gain of about 27-28 ideally, and recall I trade off another -6db for non balanced so i am back down at 21db again with 476ohms...238 is a value I happen to have a few spares of

[Joe Malone]

Well the bad news is on power up I have a 25v positive to 25v negative voltage swing on all outputs on every speaker output to ground. Some of them appear to swing back and forth with no inputs connected or loads...

Edit...actually they all swing...but all play absolutely fine when they have an audio source connected.

Update .... Well if I connect a source on the input then the output works for that channel and is fine so I assume its going into massive oscillation on the unterminated inputs. Never seen this on my other 8 channel build. Only difference is I used wima 0.1 caps on the bypass caps and 1k resistors for the gain setting. And this build is 25-0-25 where as the other one is 18-0-18 AC as I am not bridging any channels, this build is unbalanced inputs only and 478ohms for the gain. Pin 3 is grounded at the inputs for all channels.

Joe is there a need to have the trim pots installed and fully open for solving the oscillation issue or just a resistor of some value across the inputs or on the boards...if so where would i put the resistor and what value?

Any thoughts anyone?

Have you got the RCA inputs wired RCA shield to -in and RCA center +in with a 0v shield only connected at the amp end and not connected to the RCA end ? If bridged second amp has -in and +in reversed to the RCA. Ideally the RCA sockets would be insulated types that do not touch the chassis.

Be careful that you keep inputs and outputs away from each other or that can cause oscillation when nothing is plugged in.

[wizardofoz]

Joe do you mean the rca end should have a single ground back to the ground plane via a seperate cable and have the ground / shield connected at the board end for each input cable?

Right now i have a single ground cable from the chassis star to all the rca grounds, and then there is no ground connection at the board end but all the cables are grounded to the shield in cable at the rca end. The - pin of the xlr input is grounded and the hot pin is to the rca hot. The rca's are unsulated from the chassis but tied back to it.

In other words i am not sure if your last comment was how it should be or you are asking if it is that way that it shouldnt be.

I think i did it differently in the other amp i built...might have to open it up and see...that one i use as 3 channels fully balanced and 2 unbalanced but all used the xlr inputs...that one has no issues but I am using the 1k gain resistors and the problem one is now ~476 ohms and about to drop to 238 ohms to get the rca input gain up to around 27-28 again.

[wizardofoz]

Well I have rewired the speaker connection to have the grounds all back to the amp board ground plane.

The rca's have all got cable grounds to amp board also. I can play signal on all channels and its all working fine.

If any of the inputs are open/floating ie no signal input then all the channels without it will oscillate at basically +/- 24-26VDC and speakers will cook.

I really cant see what is causing this...there are no input trim pots on any channel. The input impedance on each channel measures ~9.8-10K. the - inputs are all tied to grnd and the + are all to RCA.

Am I still missing something here? the only difference to my other one of these boards is this one is 8 single channels and the other is bridged for 6 of the amps (to form 3 ch @ 140W) and the other 2 are single...all of these are setup for balanced input. The all run OK when used with RCA to XLR pin + and - to grnd too. The transormer used on this one is 18-0-18 where the 8 channel is 25-0-25 as non of the channels are bridged

[Joe Malone]

OK this is how we deal with RCA input here

So when the amp is unplugged the amp is balanced properly still.
Also when plugged in it will allow balanced noise canceling to still work and stops any ground loops.

We did some amps recently with the below setup for best gain using RCA input above
We link out R3 & R4 with zero ohm resistors and replace the trim pot with a 10k fixed resistor.
Then if more gain needed lower 1k R11 value.
Only one power ground 0V with +V and -V to 8PAKAMP PCB.
You can shield the RCA input wiring by using balanced cable but only connect 0v ground at 8PAKAMP PCB but not to RCA.

Hopefully that makes things clearer. Otherwise if still having problems post or email me some recent photos of the PCB and wiring.

Got your message about 8PAKAMP Schematic. Just need to drop me a email so I can send you one as I cannot find your email address.

[wizardofoz]

That makes a lot more sense now. I'm not sure about the R3&4 (after the input cd blocking cap) how much gain would that give on its own with them as links (0ohm).

At this time I have ~170 ohms instead of the R11 1K on the diagram. I recall that you mentioned 470 ohm would give +6db and 220 around another +6db total +12db, but dont I lose 6db with an unbal input? So the net of just R11 @ 170ohms with unbal gives ~+8db gain (I'm guessing) on the original 21db? I assume one mustn't go too low here (R11). Maybe I am too low already?

How much more would say 5K or 1K or 0 ohms on the R3/4 combo would give me and would it be better to reduce these than the 1K R11? Or should one do all?

What impact does the 10K pot have if its close to max? I would only use that to balance the channels as its an active system that expects everything the same gain. You mentioned using a 10 fixed resistor in place of the pot.

http://www.linkwitzlab.com/LX521/Description.htm is what I am using this for...at least I am hoping it will be enough... if not for the LX's than it should be OK for the Orions which are a little less demanding the bass department.

Sorry for asking more questions but I hope it makes things clearer too for others building this fine kit...it really is a nice compact design and seems to be pretty robust despite the trails I have had with this build...its still all working. Credit to the quality of the board as it has had several cable changes and surviving quite well with desoldering etc.

[Joe Malone]

That makes a lot more sense now. I'm not sure about the R3&4 (after the input cd blocking cap) how much gain would that give on its own with them as links (0ohm).

Gives a extra 12dB of gain. We found this to be the best way to get the extra gain before making the 1k R11 smaller in value as it keep the noise floor much lower.
This how the new 12PAKAMP PCB works. The trim pot on it is the R11 resistor not a input pad.

What impact does the 10K pot have if its close to max? I would only use that to balance the channels as its an active system that expects everything the same gain. You mentioned using a 10 fixed resistor in place of the pot.

It will help keep the amp gain balanced and stable when unplugged when used with the RCA wiring I showed in the last post. Just needs to be a 10k fixed resistor or the pot left open at full 10k as it will only lower the input resistance not the level if you turn it down.

http://www.linkwitzlab.com/LX521/Description.htm is what I am using this for...at least I am hoping it will be enough... if not for the LX's than it should be OK for the Orions which are a little less demanding the bass department.

We are using huge dipole system with 20 of those woofers with 12 x one metre ribbons for 1000 seater theater here at the moment using our new 12PAKAMP PCB that plugs direct into a Hypex DLCP DSP Card to do all the correction. This makes a minimal wiring solution. Uses 60 Amps to make 30 bridged 140W amps. See our jlmaudio instagram photos and facebook pages for images.

Sorry for asking more questions but I hope it makes things clearer too for others building this fine kit...it really is a nice compact design and seems to be pretty robust despite the trails I have had with this build...its still all working. Credit to the quality of the board as it has had several cable changes and surviving quite well with desoldering etc.

Link out the 2 resistors as the previous post mentioned and you should only at worst case need to make R11 470R. But always it is preferable to have a preamp that can output a better level from in to out to keep signal to noise at it best. We found we had heaps of level with on the 2 resistor linked out. With pro level systems putting out +24dBM easy we usually have to attenuate the input signal which is why the 8PAKAMP is setup with a variable input pad. The new 12PAKAMP is feed by the 2.3v output of the DLCP so needed to be able gain the amp so it clipped at the same point. We found if the amp was kept fully balanced in it with our unique circuit it could be run at less than the minimum 10x gain needed for stability and be perfectly stable with no bandwidth caps needed at all.

Really should be looking at why the preamp/xover design cannot put out enough level into the amp set with R11 as 1k to be able to drive it into clipping. Any good preamp should be able to do this easy.

[wizardofoz]

12PAKAMP another teaser? Tell us more

[Joe Malone]

12PAKAMP another teaser? Tell us more

As always the latest workshop happening are at the links below

http://www.facebook.com/jlmaudio1

http://instagram.com/jlmaudio

https://twitter.com/jlmaudio

If on instagram please hashtag any photos of JLM audio gear and kits with jlmaudio so we can find them easily

[berggren]

Hi Joe,

I have just finished the 8 Channel amp and I am happy with the results.
I use a miniDSP 4x10HD as a preamplifier. The idea for miniDSP came from LinkwitzLab.com. Linkwitz uses the miniDSP for his active loudspeakers. I have done the same and the results are indeed as Linkwitz describes on his website: breathtaking. Because of the miniDSP + 8 channel amp I can now remove 3 stereo amps and a preamp from my system which saves a lot of space.
The miniDSP has 8 channels with cross over filters and parametric equalizers on each channel. In addition to individual level controls for each channel it has one common volume control for all 8 channels which can be operated with a remote control. It has analog and digital inputs.

The cabinet for the amplifier was bought on ebay in China. I have used a 500W 2x24V toroidal trafo and and 2xLT1083 to obtain +/- 30V.
I have added a soft start unit to avoid a current peak while switching on.

Thanks again for a wonderful product.

Wout Berggren, The Netherlands.

[Joe Malone]

Hi Joe,

I have just finished the 8 Channel amp and I am happy with the results.
I use a miniDSP 4x10HD as a preamplifier. The idea for miniDSP came from LinkwitzLab.com. Linkwitz uses the miniDSP for his active loudspeakers. I have done the same and the results are indeed as Linkwitz describes on his website: breathtaking. Because of the miniDSP + 8 channel amp I can now remove 3 stereo amps and a preamp from my system which saves a lot of space.
The miniDSP has 8 channels with cross over filters and parametric equalizers on each channel. In addition to individual level controls for each channel it has one common volume control for all 8 channels which can be operated with a remote control. It has analog and digital inputs.

The cabinet for the amplifier was bought on ebay in China. I have used a 500W 2x24V toroidal trafo and and 2xLT1083 to obtain +/- 30V.
I have added a soft start unit to avoid a current peak while switching on.

Thanks again for a wonderful product.

Wout Berggren, The Netherlands.

Wow fantastic finished amp. Great work.

[atiska1]

I ended up using 476ohms (2x238 in series) but im using un balanced so might have to drop that to maybe 238 to get back up to around 27-28db gain. Will have to measure it once I am done.

Much the same as my 5 channel version but with a 25-0-25 500va as no bridging going on here. Using 2 x 8 way Speakons for the speakers to connect to my Linkwitz Orion's

Excellent. I really like the layout of your amps. You will find that you will get better performance (lower THD hum and noise etc) if we take the 0v -speaker wire from the provided pad holes on each amp on the PCB ground plane. This is due to the star ground being before the capacitor bank which is on the 8PAKAMP PCB. So the current in the 0v power wire going to the PCB modulates. So the smoothest best place to take the 0v is on the PCB where the cap bank 0v ground plane is. This would make your star ground a lot smaller as well.

About to use a lot of these 8PAKAMP'S in a new 1000 seater theater for a full dipole speaker system that is huge.

There are 2 options on the PCB to connect the 0v. I marked them with black and red. Which one shall I connect 0v -speaker wire to?

[Joe Malone]

I ended up using 476ohms (2x238 in series) but im using un balanced so might have to drop that to maybe 238 to get back up to around 27-28db gain. Will have to measure it once I am done.

Much the same as my 5 channel version but with a 25-0-25 500va as no bridging going on here. Using 2 x 8 way Speakons for the speakers to connect to my Linkwitz Orion's

Excellent. I really like the layout of your amps. You will find that you will get better performance (lower THD hum and noise etc) if we take the 0v -speaker wire from the provided pad holes on each amp on the PCB ground plane. This is due to the star ground being before the capacitor bank which is on the 8PAKAMP PCB. So the current in the 0v power wire going to the PCB modulates. So the smoothest best place to take the 0v is on the PCB where the cap bank 0v ground plane is. This would make your star ground a lot smaller as well.

About to use a lot of these 8PAKAMP'S in a new 1000 seater theater for a full dipole speaker system that is huge.

There are 2 options on the PCB to connect the 0v. I marked them with black and red. Which one shall I connect 0v -speaker wire to?

The red one closest to the amp output pad for each amp is the one to use for -output . Use one or two of the closest to centre black pads for power 0v.