Hi Mr Feedback;
>>>...Do TDA2030 need to be insulated from the heatsinks ???...<<<
If the heatsinks are grounded and you're using a split supply, yes. Otherwise, no (check the chip's app guide and you'll find suggestions for using it either with split or single supplies). I was using a split supply and grey plastic insulators (and checked with a DMM - according to it they were electrically insulated).
The new layout will probably use a heatsink INSIDE the chassis that will be left floating, so (I hope) there won't be any problems there.
Frankly the thing that got me interested in chipamps was saving money over my traditional valves, since some changes in my family situation are really cutting away at discretionary funds. Now I'm just plain getting stubborn about getting it right, so I'm going to keep grinding away at the problems until I have my building techniques down.
Ja mata,
Morse
>>>...Do TDA2030 need to be insulated from the heatsinks ???...<<<
If the heatsinks are grounded and you're using a split supply, yes. Otherwise, no (check the chip's app guide and you'll find suggestions for using it either with split or single supplies). I was using a split supply and grey plastic insulators (and checked with a DMM - according to it they were electrically insulated).
The new layout will probably use a heatsink INSIDE the chassis that will be left floating, so (I hope) there won't be any problems there.
Frankly the thing that got me interested in chipamps was saving money over my traditional valves, since some changes in my family situation are really cutting away at discretionary funds. Now I'm just plain getting stubborn about getting it right, so I'm going to keep grinding away at the problems until I have my building techniques down.
Ja mata,
Morse
If the devices are insulated from the heatsink then it should be fine to ground the heatsink directly.
In my experience, running seperated power feeds and seperated grounding connections is mandatory for stability with these amplifiers.
Can you give a circuit diagram and layout diagram of your problem circuit ?.
Eric.
In my experience, running seperated power feeds and seperated grounding connections is mandatory for stability with these amplifiers.
Can you give a circuit diagram and layout diagram of your problem circuit ?.
Eric.
Thanks Eric;
>>>...In my experience, running seperated power feeds and seperated grounding connections is mandatory for stability with these amplifiers...<<<
Interesting. I am using discrete ground shackles, though the power supply was laid out similarly to the way I would for a valve amp - in other words, I have the transformer feeding a PS board (with 4 Fast Response diodes makig a bridge, 2x 2200uF caps in series, and 2x.1uF caps in series; the caps are wired conventionally to form a split supply, and a pair of bleeder resistors are also used). Off the board are 5 wires, 2 Vpositive, 2 Vnegative, and a single ground shackle to the star ground on the chassis (by the IEC inlet). Each chip has a pair of .1uF miniature ceramic disc caps that are tied together with a common lead to ground. Of course one "free" leg of one cap goes to the Vpositive pin on the chip and the "free" leg of the other cap goes to the Vnegative pin on the chip. The ground point in between the two caps ties to ground with a discrete ground shackle.
>>>...Can you give a circuit diagram and layout diagram of your problem circuit ?...<<<
Sorry I don't have any graphics tools that would let me draw it out easily! Hopefully the text description above is adequate.
The exact schemo I used was posted earlier in this thread (it's the manufacturer's recommended split supply configuration for the TDA2030A) by someone else.
Before completely giving up on this configuration, I have considered adding a pair of 100uF caps to each chip (as indicated on the schemo) as close to the chip pins as possible.
However, it's beginning to sound like my use of a pair of common filter caps earlier in the circuit is a no-no with these devices. Is that accurate, or am I misreading your post?
Can I use a common diode bridge, and just use separate filters for each chip?
Thanks again for any help or suggestions!
All the best,
Morse
>>>...In my experience, running seperated power feeds and seperated grounding connections is mandatory for stability with these amplifiers...<<<
Interesting. I am using discrete ground shackles, though the power supply was laid out similarly to the way I would for a valve amp - in other words, I have the transformer feeding a PS board (with 4 Fast Response diodes makig a bridge, 2x 2200uF caps in series, and 2x.1uF caps in series; the caps are wired conventionally to form a split supply, and a pair of bleeder resistors are also used). Off the board are 5 wires, 2 Vpositive, 2 Vnegative, and a single ground shackle to the star ground on the chassis (by the IEC inlet). Each chip has a pair of .1uF miniature ceramic disc caps that are tied together with a common lead to ground. Of course one "free" leg of one cap goes to the Vpositive pin on the chip and the "free" leg of the other cap goes to the Vnegative pin on the chip. The ground point in between the two caps ties to ground with a discrete ground shackle.
>>>...Can you give a circuit diagram and layout diagram of your problem circuit ?...<<<
Sorry I don't have any graphics tools that would let me draw it out easily! Hopefully the text description above is adequate.
The exact schemo I used was posted earlier in this thread (it's the manufacturer's recommended split supply configuration for the TDA2030A) by someone else.
Before completely giving up on this configuration, I have considered adding a pair of 100uF caps to each chip (as indicated on the schemo) as close to the chip pins as possible.
However, it's beginning to sound like my use of a pair of common filter caps earlier in the circuit is a no-no with these devices. Is that accurate, or am I misreading your post?
Can I use a common diode bridge, and just use separate filters for each chip?
Thanks again for any help or suggestions!
All the best,
Morse
As promised, the PCBs for the 1875 amp
Just designed them today, haven't tried them yet.
Anyway, heres all the stuff, all in pdf format. hopefully it is right.
Just designed them today, haven't tried them yet.
Anyway, heres all the stuff, all in pdf format. hopefully it is right.
Doing It Right........
If this is your circuit, you need to keep the earth connections of R4 and C1 (electrically) right away from each other, and away from C4//C6 and C3//C5 grounds.
The Zobel network is best across the speaker output terminals.
Understand that if the ground of R4 is modulated by speaker/psu currents, this constitutes positive feedback and will encourage oscillation.
If the ground of C1 is modulated by speaker/psu currents this further contributes intermodulation and maybe instability.
It is ok to run a single psu module, but the grounds of these R4, C1, C4//C6 and C3//C5 components require their own clean connections back to psu central ground.
The input ground and speaker grounds need to reference back to this psu central ground also.
Maybe this is clearer for you now.
Eric.
If this is your circuit, you need to keep the earth connections of R4 and C1 (electrically) right away from each other, and away from C4//C6 and C3//C5 grounds.
The Zobel network is best across the speaker output terminals.
Understand that if the ground of R4 is modulated by speaker/psu currents, this constitutes positive feedback and will encourage oscillation.
If the ground of C1 is modulated by speaker/psu currents this further contributes intermodulation and maybe instability.
It is ok to run a single psu module, but the grounds of these R4, C1, C4//C6 and C3//C5 components require their own clean connections back to psu central ground.
The input ground and speaker grounds need to reference back to this psu central ground also.
Maybe this is clearer for you now.
Eric.
Please accept my apologies Eric;
I'd claimed that the schemos were posted elsewhere on this thread and that was in error. It was in an earlier thread when I was having trouble getting one channel off the ground (progress of sorts in that now I have one very reliable channel).
Hopefully I've done this right and they're visible here.
Thanks in advance for any help or advice.
Ja mata,
Morse
I'd claimed that the schemos were posted elsewhere on this thread and that was in error. It was in an earlier thread when I was having trouble getting one channel off the ground (progress of sorts in that now I have one very reliable channel).
Hopefully I've done this right and they're visible here.
Thanks in advance for any help or advice.
Ja mata,
Morse
Attachments
Hi Eric;
I just read your most recent post. Okay, I can use a common PS module, but it looks like I need to put some larger decoupling caps next to each chip. Aside from 'buddying up' the grounds between V+ and V- on each chip, I think I'm following your grounding schema closely (lots of discrete ground wires, and the signal leads all ground to the chassis at the volume control at the front, though the output ground ties directly to the star ground. Perhaps I should have a separate 'substar' for it?
Doumo arigatou gozaimasu for your help!
Ja mata,
Morse
I just read your most recent post. Okay, I can use a common PS module, but it looks like I need to put some larger decoupling caps next to each chip. Aside from 'buddying up' the grounds between V+ and V- on each chip, I think I'm following your grounding schema closely (lots of discrete ground wires, and the signal leads all ground to the chassis at the volume control at the front, though the output ground ties directly to the star ground. Perhaps I should have a separate 'substar' for it?
Doumo arigatou gozaimasu for your help!
Ja mata,
Morse
Actally, my schematic is the first one on the datasheet. Basically, it is the same as the one mrfeedback posted, but with a 1M resistor before the input to ground. I guess it isn't neccesary, though.
So, am I to assume that my layout will not work well? I would be powering 2 channels off of a single transformer.
Thanks!
-Mike
So, am I to assume that my layout will not work well? I would be powering 2 channels off of a single transformer.
Thanks!
-Mike
A zobel network redirects hi-frequency oscillations to ground, if your amp is suffering because of them.
It may or may not be necessary, depends on your system
It may or may not be necessary, depends on your system
The Zobel network ensures that the amplifier is loaded at high frequencies.soundNERD said:
It also provides a loading to induced external pickup in the speaker leads.
Eric.
Will it be needed for the 1875. All the schematics i've seen for it have that on it. I would like to save on parts.
-Mike
-Mike
Yes put the Zobel networks directly across the output terminals - and the component cost is minimal.
Eric.
Eric.
Hi,
Of the speaker unit I hope, not the amp?
Sorry for not having read the entire thread...
Cheers, 😉
Of the speaker unit I hope, not the amp?
Sorry for not having read the entire thread...
Cheers, 😉
Hi, everybody
I am newbei to amp
I have a car amp that was overloaded and now it's under protection mode. I ran the amp bridged at 2ohms which was not legal by this amp model. And now the amp switches to procection mode each times I try to hook it up.
How can I get the amp out of protection mode?
I am newbei to amp
I have a car amp that was overloaded and now it's under protection mode. I ran the amp bridged at 2ohms which was not legal by this amp model. And now the amp switches to procection mode each times I try to hook it up.
How can I get the amp out of protection mode?
Sounds like you blew something inside the amp.
If an amp's manual says don't use a 2-ohm speaker, then don't use a 2-ohm speaker.
What happens with lower impedences is the amps run hotter. Most likely what happened is you overheated the chips inside it.
Did you check the obvious first, such as replacing the fuses? Have you tried running it with no load (no speaker)? and seeing if the unit turns on?
What are you powering with it? I would assume 2 4-ohm subs in parallel connection. If so, connect them in series, and you will get an 8-ohm load, and therefore not blow the amp.
-Mike
If an amp's manual says don't use a 2-ohm speaker, then don't use a 2-ohm speaker.
What happens with lower impedences is the amps run hotter. Most likely what happened is you overheated the chips inside it.
Did you check the obvious first, such as replacing the fuses? Have you tried running it with no load (no speaker)? and seeing if the unit turns on?
What are you powering with it? I would assume 2 4-ohm subs in parallel connection. If so, connect them in series, and you will get an 8-ohm load, and therefore not blow the amp.
-Mike
I have finished my amp using TDA2030A for a while now, it sounds great. I shared two channels on a single power supply, +/- 18 volts. I used two 4700uF caps(kenwood)
see if this helps
see if this helps
Thanks for the tip, Skyraider (Douglas A-1H ?);
I should've kept updating this thread!
As of now, I've got both channels working without oscillation, BUT I've got a persistant hum in the left channel that's only audible through the headphone circuit (used a 150r/10r voltage divider and duplicated the zoebel on the 10r leg of it) or with your ear up against the left speaker. Took a while to figure out, but it looks like the signal lead is too long and is picking up hum off the power supply (both channels had it initially, but my right channel's signal lead was WAY longer than it had to be - trimming it cured the problem in that channel).
Now I'm looking at revising the layout a little so that both chips are up front next to the volume control pot. I may replace the 26ga signal wire from the RCA in's to the pot with shielded mike cable (pain in the *** to work with, but it keeps things quiet...).
Initially I'd thought heating would be a big issue so I stupidly put one chip on either side of the PS board to keep 'em as far apart as possible. Turns out the TDA2030A runs cold as the grave most of the time and only heats up to luke warm (with a 5" by 5-1/2" chassis as a common heat sink) after heavy use into an 8 ohm load. Going to be a pain replacing some of the leads that are too short though....
Thanks again and all the best!
Ja mata,
Morse
I should've kept updating this thread!
As of now, I've got both channels working without oscillation, BUT I've got a persistant hum in the left channel that's only audible through the headphone circuit (used a 150r/10r voltage divider and duplicated the zoebel on the 10r leg of it) or with your ear up against the left speaker. Took a while to figure out, but it looks like the signal lead is too long and is picking up hum off the power supply (both channels had it initially, but my right channel's signal lead was WAY longer than it had to be - trimming it cured the problem in that channel).
Now I'm looking at revising the layout a little so that both chips are up front next to the volume control pot. I may replace the 26ga signal wire from the RCA in's to the pot with shielded mike cable (pain in the *** to work with, but it keeps things quiet...).
Initially I'd thought heating would be a big issue so I stupidly put one chip on either side of the PS board to keep 'em as far apart as possible. Turns out the TDA2030A runs cold as the grave most of the time and only heats up to luke warm (with a 5" by 5-1/2" chassis as a common heat sink) after heavy use into an 8 ohm load. Going to be a pain replacing some of the leads that are too short though....
Thanks again and all the best!
Ja mata,
Morse
I used shielded audio cable for the inputs too. I am using a way too small heatsink for my TDA2030A's, it gets really hot, but it never blows up untill now. I dont worry much about it because those chips have thermal shutdown, but mainly because I am too LAZY to change a bigger one. it shortens the lifespan, of course, keep your chips cool and they keeps you happy=)
I copied off the PCB design that is recomended in the datasheet, it has HUGE tracks to cope with all the currents
I copied off the PCB design that is recomended in the datasheet, it has HUGE tracks to cope with all the currents
- Status
- Not open for further replies.