Im planning to use 1.5Amp bias per ALF16N20P at 35V resulting 50W of heat.
as per Owen he drove each mosfet at that similar dissipation and I have few questions.
How much heatroom do we have for audio signal. Do I need to use two ALF16N20P in that case or one is enough? like 50W per one fet or 25W for each fet at that voltage?
Is ALF16N20P is stable at 50W of continuous heat dissipation? at 35V 1.5A bias?
Pls explain why?, which means to define your signal source level and all the gain stages between it and the speakers.
If your signal source is a CD with 2Vrms output, I agree you do not need as much gain since you will always be attenuating the source level as the wire amp has a gain of 25 or 28.3dBv. With no attenuation and a 2Vrms i/p you will get 50Vrms(>70V DC supply rails) into 8 ohm or 312W. Typical power amps usually have around 1Vrms input sensitivity. The engineers knew what they are doing. If you still want less gain, than what is recommended, I believe that you can use external compensation!!
I have source of miniSHARC + DAC 2VRMS SE output into LPUHP + B&C DE250 and LME49830 + parallel 8in peerless.
I figure I only need more like 22 or 23dB gain (105dB peak output at listening position) rather than the minimum setting of 26dB. For all the effort people go to for improving performance of amps and component selection 3dB seems like a compromise.
Using IRFP240/9140N - any changes?
Hi Guys
I just ordered the Wire Amp pcb:s (thanks OPC
) and want to use IRFP240/9140N instead of the lateral fets.
Do you think it´s an ok choice for this application, or should I try to get the lateral fets instead?
Are there any component modifications that are necessary?
Thanks
Hi Guys
I just ordered the Wire Amp pcb:s (thanks OPC
) and want to use IRFP240/9140N instead of the lateral fets.Do you think it´s an ok choice for this application, or should I try to get the lateral fets instead?
Are there any component modifications that are necessary?
Thanks
You'll have to stick with lateral FETs otherwise temperature compensation will be needed for bias, and there are no provisions for that on the PCB. You will also need emitter resistors and likely a few other modifications.
I'm sure it could be made to work, but I wouldn't recommend it.
A wide variety of lateral FETs will work in the design, so you should be able to find something locally.
Regards,
Owen
I'm sure it could be made to work, but I wouldn't recommend it.
A wide variety of lateral FETs will work in the design, so you should be able to find something locally.
Regards,
Owen
OK - thanks for a quick answer! I'll go for the L-FET!
When I go through the BOM and Mouser-list that rsavas did (thanks for that) there are too many components that I should by 10 of to get a better price, when I only need 2.
I guess there are a lot of components "left over" out there.
If someone has a decent amount of parts needed to populate these board (2pcs), please send me a message.
Thanks
When I go through the BOM and Mouser-list that rsavas did (thanks for that) there are too many components that I should by 10 of to get a better price, when I only need 2.
I guess there are a lot of components "left over" out there.
If someone has a decent amount of parts needed to populate these board (2pcs), please send me a message.
Thanks
Hi guys,
Yes, I still have not put one together yet. I have gone over it again and discovered I am out of synch, nothing new
Can someone point me to the link for the V2 pcb schematic, pls
The schematic(s) that I have (09-13-2011, Rev A,drawing 01-04) are different than the V2 pcb, re: R40,48.
On the V2 pcb R48, C76(+) connect to (-) input.
I also plan to do a few mods, to increase the gain for 31dB, I will use SE operation, change the input series & low side feedback R's to 499, input R 15K to gnd, and the feedback R from the o/p. Add a 1nF across the 15K pulldown. Input coupler to a Wima 4.7uF film (505-MKS2-4.7/50/10). Add a zobel to the output 3.9/47nF. I do not need the 0.01% R's ( preferred for balanced op.) so found the cheaper 0.05% thin film R (754-RR1220P-4990D-M)
Thx
Rick
Yes, I still have not put one together yet. I have gone over it again and discovered I am out of synch, nothing new
Can someone point me to the link for the V2 pcb schematic, pls
The schematic(s) that I have (09-13-2011, Rev A,drawing 01-04) are different than the V2 pcb, re: R40,48.
On the V2 pcb R48, C76(+) connect to (-) input.
I also plan to do a few mods, to increase the gain for 31dB, I will use SE operation, change the input series & low side feedback R's to 499, input R 15K to gnd, and the feedback R from the o/p. Add a 1nF across the 15K pulldown. Input coupler to a Wima 4.7uF film (505-MKS2-4.7/50/10). Add a zobel to the output 3.9/47nF. I do not need the 0.01% R's ( preferred for balanced op.) so found the cheaper 0.05% thin film R (754-RR1220P-4990D-M)
Thx
Rick
If you take another look, I have updated the Mouser BOM for my application. I left on the original R's that OPC specified and added a few extras for other projects.
I recall I made a mistake on one of the big ecaps, I think I originally had a 5mm lead spaces where the pcb is 7.5mm LS (1000uF/100V). So I used as a sub for the 647-UPW2A102MHD, a
661-EKYB101E102MM40S
I bumped the qty to multiples of 10, for the cheap passives, but you can modify for your own liking too.
http://www.mouser.com/ProjectManager/ProjectDetail.aspx?AccessID=d36806234b
I recall I made a mistake on one of the big ecaps, I think I originally had a 5mm lead spaces where the pcb is 7.5mm LS (1000uF/100V). So I used as a sub for the 647-UPW2A102MHD, a
661-EKYB101E102MM40S
I bumped the qty to multiples of 10, for the cheap passives, but you can modify for your own liking too.
http://www.mouser.com/ProjectManager/ProjectDetail.aspx?AccessID=d36806234b
Sorry have to post again, I timed out on the edit.
If you take another look, I have updated the Mouser BOM for my application. I left on the original R's that OPC specified and added a few extras for other projects.
I recall I made a mistake on one of the big ecaps, I think I originally had a 5mm lead spaces where the pcb is 7.5mm LS (1000uF/100V). So I used as a sub for the 647-UPW2A102MHD, a
661-EKYB101E102MM40S
I bumped the qty to multiples of 10, for the cheap passives, but you can modify for your own liking too.
http://www.mouser.com/ProjectManager/ProjectDetail.aspx?AccessID=d36806234b
Sorry that the mouser BOM is a mixed bag of qty, for example, the LME49830 I show as only 1, because I already have 4 of them and I have 4 boards in hand.
When/if you go to order from my mouser BOM, you are free to add/delete/change for your liking to cost reduce, so if you only need one R, then just change your quantity.
Note I have not built to this BOM yet, so do a sanity check pls, I could have made a mistake, if so pls let me know and I'll correct it.
If you take another look, I have updated the Mouser BOM for my application. I left on the original R's that OPC specified and added a few extras for other projects.
I recall I made a mistake on one of the big ecaps, I think I originally had a 5mm lead spaces where the pcb is 7.5mm LS (1000uF/100V). So I used as a sub for the 647-UPW2A102MHD, a
661-EKYB101E102MM40S
I bumped the qty to multiples of 10, for the cheap passives, but you can modify for your own liking too.
http://www.mouser.com/ProjectManager/ProjectDetail.aspx?AccessID=d36806234b
Sorry that the mouser BOM is a mixed bag of qty, for example, the LME49830 I show as only 1, because I already have 4 of them and I have 4 boards in hand.
When/if you go to order from my mouser BOM, you are free to add/delete/change for your liking to cost reduce, so if you only need one R, then just change your quantity.
Note I have not built to this BOM yet, so do a sanity check pls, I could have made a mistake, if so pls let me know and I'll correct it.
I found 2 schematics. One on the first page and the other on page 75...post 742.
I edited rsavas Mouser project to be just components to build a SE-version (AC or DC).
http://www.mouser.com/ProjectManager/ProjectDetail.aspx?AccessID=31B5B2FF32
@rsavas
It seems you picked 3 different isolators for TO247 - which is most suitable?
1. Kapton, 0,15mm, 6kV
2. Silicone, 0,23mm, 5kV
3. Alu oxide, 2mm,
Is the heatsink for LME big enough? On the picture it seems smaller than the one OPC picked from Digikey.
I find 2 LATFETs, Exicon (10N20, 10P20) and Renesas(2SK1058,2SJ162) - which one should I go for? Do they need isolators?
Thanks!
http://www.mouser.com/ProjectManager/ProjectDetail.aspx?AccessID=31B5B2FF32
@rsavas
It seems you picked 3 different isolators for TO247 - which is most suitable?
1. Kapton, 0,15mm, 6kV
2. Silicone, 0,23mm, 5kV
3. Alu oxide, 2mm,
Is the heatsink for LME big enough? On the picture it seems smaller than the one OPC picked from Digikey.
I find 2 LATFETs, Exicon (10N20, 10P20) and Renesas(2SK1058,2SJ162) - which one should I go for? Do they need isolators?
Thanks!
Last edited:
The only v2 I'm aware of is Owen's PSU PCB. The notes from his for sale thread are:
I used assembly manual and schematic from post 742 when assembling mine. I'm not sure what you're asking for with the 'v2' PCB. The change from green to white PCBs includes different routing but otherwise same circuit is my understanding.
Chris
Hi guys,
Yes I know it is the same circuit but OPC changed reference designations, so far I found R40/R48 reversed, I do not of any other's yet. My pcb is the white one, I assume is from the second rev or batch. IIRC OPC added a forth layer on the V2.
I am looking for a change log or a schematic for V2!!
As far as insulators goes, OPC did not spec any on the BOM, so we need insulators for 2 TO-247 latfet's and one for the LME49830. I chosoe a few to test out. Until I do an assembly/test, I will not know for sure what is the best to recommend.
People who have assembled should respond to what material they used for there assembly.
Neither of these, would I recommend.
Yes they need insulators, the metal flange are live HI Voltage, caution is very much needed. Yes it can kill you or give you a good shock.
Do not over torque the mounting of these devices. Always do a ohmmeter, check to see that the flange is isolated from ground. The Heatsink is usually at ground/chassis potential. Suggest to do a thourough reading of the assembly documentation again.
Use the recommended Lafets, in your first attempt, as was posted earlier by OPC. I could not get ALF16P20W and they had a sale on ALF16P160W/ALF16N160W, so I bought them instead. My supply is not to be over 70V, so the rules was Vceo max > (2 * Vcc)
If you want to play around with other types later, great go for it.
P-Channel
“Pro Audio” ACD103PDD
“Semelab” ALF16P20W
“Exicon” ECW20P20Z
“Magnatec” BUZ906DP
“Magnatec” BUZ907DP – higher voltage
“Magnatec” BUZ908DP – higher voltage
N-Channel
“Pro Audio” ACD101NDD
“Semelab” ALF16N20W
“Exicon” ECW20N20Z
“Magnatec” BUZ901DP
“Magnatec” BUZ902DP – higher voltage
“Magnatec” BUZ903DP – higher voltage
Yes I know it is the same circuit but OPC changed reference designations, so far I found R40/R48 reversed, I do not of any other's yet. My pcb is the white one, I assume is from the second rev or batch. IIRC OPC added a forth layer on the V2.
I am looking for a change log or a schematic for V2!!
As far as insulators goes, OPC did not spec any on the BOM, so we need insulators for 2 TO-247 latfet's and one for the LME49830. I chosoe a few to test out. Until I do an assembly/test, I will not know for sure what is the best to recommend.
People who have assembled should respond to what material they used for there assembly.
Neither of these, would I recommend.
Yes they need insulators, the metal flange are live HI Voltage, caution is very much needed. Yes it can kill you or give you a good shock.
Do not over torque the mounting of these devices. Always do a ohmmeter, check to see that the flange is isolated from ground. The Heatsink is usually at ground/chassis potential. Suggest to do a thourough reading of the assembly documentation again.
It is what is spec'd and what I will use. I know AndrewT thought that it was too small, but you can add a secondary HS to the existing one, this is what I plan to do, if I find it is not adequate.
Use the recommended Lafets, in your first attempt, as was posted earlier by OPC. I could not get ALF16P20W and they had a sale on ALF16P160W/ALF16N160W, so I bought them instead. My supply is not to be over 70V, so the rules was Vceo max > (2 * Vcc)
If you want to play around with other types later, great go for it.
P-Channel
“Pro Audio” ACD103PDD
“Semelab” ALF16P20W
“Exicon” ECW20P20Z
“Magnatec” BUZ906DP
“Magnatec” BUZ907DP – higher voltage
“Magnatec” BUZ908DP – higher voltage
N-Channel
“Pro Audio” ACD101NDD
“Semelab” ALF16N20W
“Exicon” ECW20N20Z
“Magnatec” BUZ901DP
“Magnatec” BUZ902DP – higher voltage
“Magnatec” BUZ903DP – higher voltage
Sorry, the file you have requested does not exist.
For lower supply rail voltages the specified/supplied sink is good. say <=+-40Vdc
For +-50Vdc the chip runs hotter than I would like, but it will not overheat if the output is not overloaded. The adjacent capacitors get heated by the hot sink !
For >+-65Vdc, I consider the specified sink to be too small. It will run far too hot for my temperature targets.
But it not just the sink, it's the lack of space around the warm/hot sink, that exacerbates the problem at the higher supply voltages.
Last edited:
After a couple of years I'm now just about to throw together a pair of amps with these LME49830.
I have a 1.9°C/W heatsink for each LME chip and plan to use them with no silicon pad, just grease, possibly Arctic Silver as I have some, so as to get more heat out of the LME as I'm going to use as much as ±100V rails for the chip.
I found that -ve supply pin 10 is joined to the exposed metal heat pad on the rear of the chip.
With no insulator, is there a possibility of noise or interference of any kind from the heatsink going into the chip? The sink will be isolated from everything else but will have on it the regulators or pass transistors on silicone pads.
Or I could use one of those ceramic 15°C/W insulators and some grease each side and possibly earth the heat sink.
Thanks.
Edit: I just saw the previous post and thought I should point out, before anyone asks, 'how did you fit that on there', that I'm not using Owens boards. It's 100mm wide x 50mm x 40mm comb section. 10mm base thickness. Farnell 4621566 aka Fischer Electronic SK 92/50 SA
I have a 1.9°C/W heatsink for each LME chip and plan to use them with no silicon pad, just grease, possibly Arctic Silver as I have some, so as to get more heat out of the LME as I'm going to use as much as ±100V rails for the chip.
I found that -ve supply pin 10 is joined to the exposed metal heat pad on the rear of the chip.
With no insulator, is there a possibility of noise or interference of any kind from the heatsink going into the chip? The sink will be isolated from everything else but will have on it the regulators or pass transistors on silicone pads.
Or I could use one of those ceramic 15°C/W insulators and some grease each side and possibly earth the heat sink.
Thanks.
Edit: I just saw the previous post and thought I should point out, before anyone asks, 'how did you fit that on there', that I'm not using Owens boards. It's 100mm wide x 50mm x 40mm comb section. 10mm base thickness. Farnell 4621566 aka Fischer Electronic SK 92/50 SA
Last edited:
Yes, there is.
The Chip likes it sink to be grounded and electrically isolated from itself.
IME. I use thin kapton tape for isolation.