Thanks. I'd to that but, given my lack of spice fluency, I think it will take me less time to just experiment with temperature compensation.
Sheldon
Member
Joined 2009
Paid Member
Hi Gareth
TGM5 prototype PCB you've made is simply gorgeous, I'm impressed how good this toner transfer method is, excellent for DIY-ers. Now you are on the fast lane to make a very good amplifier. I looked over your schematic again and I don't see any problematical spots, only you will have to play a little with Miller compensating capacitors. I hardly wait to hear the news about results.
Regards, Andrej
Good to know the schematic is approved, gives me confidence to build all 3 boards.
I took a look at my soldering iron and realized I stand no chance of soldering the SOT 363 part with a 40W iron that has a tip larger than the part itself. Luckily, I know a kind lady who does this kind of work.
Attachments
Hi Gareth
Lovely Trio, like my speakers are hehe
Look, you are now in the phase of assembly and it would be nice not to solder input filter & compensating pF caps yet. I'm interested what's the phase shift, output impedance and gain-bandwidth limit of fully uncompensated TGM5. I presume that all these tiny smd's are high fT types so you can get similar results like I did at mine last SSA measurements. Do you have signal generator plus scope to enable you phase shift measurement at 2 MHz to have direct comparison?
The PCB is small, all traces very short - HF kind (low C, low L), almost micro-wave ready, so the layout itself enables high gain-bandwidth. The only question is if the parts are capable of doing it the same. I know you'll later restrict all parameters within safe margins but for the curiosity sakes please measure one channel extensively. Thanks.
Regards, Andrej
Lovely Trio, like my speakers are hehe
Look, you are now in the phase of assembly and it would be nice not to solder input filter & compensating pF caps yet. I'm interested what's the phase shift, output impedance and gain-bandwidth limit of fully uncompensated TGM5. I presume that all these tiny smd's are high fT types so you can get similar results like I did at mine last SSA measurements. Do you have signal generator plus scope to enable you phase shift measurement at 2 MHz to have direct comparison?
The PCB is small, all traces very short - HF kind (low C, low L), almost micro-wave ready, so the layout itself enables high gain-bandwidth. The only question is if the parts are capable of doing it the same. I know you'll later restrict all parameters within safe margins but for the curiosity sakes please measure one channel extensively. Thanks.
Regards, Andrej
Member
Joined 2009
Paid Member
Well, I decided that I need to use the soldering station at my work which is set up for SMD parts rather than what I have at home. I don't plan to invest in SMD equipment at home - the TGM5 will be my only SMD amp I think, and actually if it sounds as good as I expect it may also be my last Class AB solid state amp - I think this design is that promising!
Trouble is, we are moving buildings at work this week and next week I have an unplanned business trip next week that will slow me down a little - but this one should be the last trip this year. I am hoping to hear music before Christmas !!
In the background I am working towards preparing a chasis and I have a test bench with regulated power supplies, also a signal generator and 2 CRO's at the ready...
Trouble is, we are moving buildings at work this week and next week I have an unplanned business trip next week that will slow me down a little - but this one should be the last trip this year. I am hoping to hear music before Christmas !!
In the background I am working towards preparing a chasis and I have a test bench with regulated power supplies, also a signal generator and 2 CRO's at the ready...
Last edited:
Member
Joined 2009
Paid Member
I'm back from travel.
The smallest SMD parts are now soldered to the boards, the Quickfit connectors and solder pins for the input signal also.
The heatsink has been drilled and 'tapped'. Photo shows the centre channel board (power devices are sandwiched between pcb and heatsink but not yet soldered to the board).
A chasis is being prepared, four large EI transformers have been secured to the basic chasis. It weighs an awful lot.
The smallest SMD parts are now soldered to the boards, the Quickfit connectors and solder pins for the input signal also.
The heatsink has been drilled and 'tapped'. Photo shows the centre channel board (power devices are sandwiched between pcb and heatsink but not yet soldered to the board).
A chasis is being prepared, four large EI transformers have been secured to the basic chasis. It weighs an awful lot.
Attachments
As you probably already noticed in the meantime, I started with assembly of my SSA BIGBT two channel amp at the begining, to evaluate its sonic frontiers as stereo amp and if proved sucessful, I'll produce more channels for my main system. Hopefully you're also close to TGM5 testing phase, starting with extensive measurements.
Member
Joined 2009
Paid Member
Well, the day job left me alone today so I sat down with soldering iron to install the rest of the SMD parts. Took quite a long time even though it wasn't too difficult.
Some resistances don't look right - meaning probing between + and - power rails for example.
I will give it a good look over and only then, if I'm happy with it, I'll hook up a current-limited bench supply.
Some resistances don't look right - meaning probing between + and - power rails for example.
I will give it a good look over and only then, if I'm happy with it, I'll hook up a current-limited bench supply.
Attachments
Member
Joined 2009
Paid Member
Now the real work begins.
Both of my bench supplies are inadequate for the job - voltage too low. I have access to a better supply at work so I've started the process there (after hours).
First power
-up shows something isn't right. The front end is the issue. There is a big voltage drop across the resistors providing supply to the zener diodes - the zeners don't get a chance to set a regulated voltage at the base of the cascode devices so the excess current is flowing somewhere... It's not symmetrical but both sides have the same issue. To me, it behaves as if there is excess current (it would have to be several mA) flowing into the base(s) of the cascode devices and thereafter through the input pair. The input to the amp is floating.
fyi - I have deviated from the component values I posted earlier. I've introduced 1k degeneration for each VAS devices along with 1k5 as collector load for each cascode device. The input is shunted to ground through 10k instead of original planned 20k. No fuses fitted to the power device collectors yet - I've installed 100R 1W resistors there initially - a precaution I learned from AKSA.
Both of my bench supplies are inadequate for the job - voltage too low. I have access to a better supply at work so I've started the process there (after hours).
First power
-up shows something isn't right. The front end is the issue. There is a big voltage drop across the resistors providing supply to the zener diodes - the zeners don't get a chance to set a regulated voltage at the base of the cascode devices so the excess current is flowing somewhere... It's not symmetrical but both sides have the same issue. To me, it behaves as if there is excess current (it would have to be several mA) flowing into the base(s) of the cascode devices and thereafter through the input pair. The input to the amp is floating.
fyi - I have deviated from the component values I posted earlier. I've introduced 1k degeneration for each VAS devices along with 1k5 as collector load for each cascode device. The input is shunted to ground through 10k instead of original planned 20k. No fuses fitted to the power device collectors yet - I've installed 100R 1W resistors there initially - a precaution I learned from AKSA.
Last edited:
It looks you've mixed some SOIC components polarity - pins.
+/- zener potential should be present if everything is connected properly according to the schematic.
Member
Joined 2009
Paid Member
I checked pin-outs and part placement-orientation and see no issues. I measured voltages everywhere and from that I was able to look at current flows. The zener diodes don't get a chance to do their jobs properly when they are being starved of current. I followed up with what this was telling me with some resistance measurements and confirmed a very low resistance where one should not exist - between the emitter of one input device and collector of the other. I have my suspects of the cause of this - most likely my poor soldering, which I'll check out tomorrow.
It's encouraging to see the fish struggle on the line a little
p.s. screws are insulated with nylon washers (not ideal as they will not hold the compressive force of the screw consistently over time, fibre washers would be better)
It's encouraging to see the fish struggle on the line a little
p.s. screws are insulated with nylon washers (not ideal as they will not hold the compressive force of the screw consistently over time, fibre washers would be better)
Last edited:
It's encouraging to see the fish struggle on the line a little
Yes exactly, that's a charm of it. More struggle with the "fish", the more fun and a lot of satisfaction from the "catch".
Yes exactly, that's a charm of it. More struggle with the "fish", the more fun and a lot of satisfaction from the "catch".
The words of a cat about the fate of fish. Well, what more to expect?
Working right out of the box has his charm too ;-) And is more wife compliant.More struggle with the "fish", the more fun and a lot of satisfaction from the "catch".
Member
Joined 2009
Paid Member
The short was readily identified and rectified, no harm done. The zener diodes are now holding a nice stable voltage. Had I checked for shorts first time around it might have been a 'straight out of the box' experience so I'm quite satisfied.
I've installed the on-board 4A fuses to the power device collectors - they solder straight on top of the 100R resistors I had there before.
The dc offset at the output appears stable and negligible (after adjustment), idle current has been set at 88mA or so and I'm going to see how stable it is (idle current looks stable over a 20 minute window). The board and components are nice and happy sitting a little above room temperature.
I have a power supply at my work, but no source or speakers. I have a source and speakers at home but no power supply. I may simply bring the mountain to ..
The fish is now in the net - but not yet in the boat...
I've installed the on-board 4A fuses to the power device collectors - they solder straight on top of the 100R resistors I had there before.
The dc offset at the output appears stable and negligible (after adjustment), idle current has been set at 88mA or so and I'm going to see how stable it is (idle current looks stable over a 20 minute window). The board and components are nice and happy sitting a little above room temperature.
I have a power supply at my work, but no source or speakers. I have a source and speakers at home but no power supply. I may simply bring the mountain to ..
The fish is now in the net - but not yet in the boat...
Last edited:
Member
Joined 2009
Paid Member
She sings very nicely indeed - I hooked up a 50k volume pot, using Magnum Dynalab FM tuner source, playing through my own design MARTELLO speaker.
Attachments
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.