
Home  Forums  Rules  Articles  The diyAudio Store  Gallery  Blogs  Register  Donations  FAQ  Calendar  Search  Today's Posts  Mark Forums Read  Search 
Solid State Talk all about solid state amplification. 

Please consider donating to help us continue to serve you.
Ads on/off / Custom Title / More PMs / More album space / Advanced printing & mass image saving 

Thread Tools  Search this Thread 
27th January 2007, 09:09 PM  #1571  
diyAudio Member
Join Date: Sep 2001
Location: Pickering, Canada

Quote:
The Jordan Jx53 voice coil fried....
__________________
The Butcher 

27th January 2007, 09:15 PM  #1572 
diyAudio Member

commiserations
.... sorry to hear that  so I guess you will have to replace both ?

28th January 2007, 12:31 AM  #1573 
diyAudio Member
Join Date: Jun 2001
Location: Calgary

Really sorry to hear about your amp and speaker, Chris. Do you think it was overheating, then? I'm seriously thinking about adding an overtemperature shutdown to my amp. I don't think there's any way to guarantee saving the speakers if a transistor shorts.
For those of you who weren't following Chris's progress back a few years ago, here's a good shot of what his amp looked like: http://www.diyaudio.com/forums/attac...amp=1039327729 Weren't there casters on it too, or was that somebody else's? 
28th January 2007, 01:36 AM  #1574 
diyAudio Member

a fuse on the o/p made from a different grade of resistance wire is a quick and easy solution with very little impact on the sonics

28th January 2007, 02:34 AM  #1575  
diyAudio Member
Join Date: Sep 2001
Location: Pickering, Canada

Thanks guys
Quote:
The picture of that amp in your link is still working fine today. That is the pair of Geoff's original version of JLH for ESL using Mj15003 to power the pair of Jx150 and they have Vellemen K4700 speaker protection kits installed and regulate PSU. I have not make any changes to those. The one that fried last weekend is the newer 2 CCS and using big CLC remote power supply using Mj21194 output transistors. In my system I am using active 3 way, for each channel I am using the following 1 JLH using Mj21194 to power the Jx53 and a ribbon tweeter 1 JLH using Mj15003 to power the Jx150 2 BrianGT GC to power 2 X 12 inche H frame subwoofers This time the Mj21194 burned out due to second breakdown means overheat over a prolonged period shorten the life span of these transistors. Luckyly the ribbon has a transformer to survive this ordeal but the Jx53 is not that lucky. Mike, I have to try and find one Jx53 only for now otherwise I have to find a new speaker project suitable for OB three way. Now it is very difficult to go back to boxes since I have acquired the taste of open baffle. Can you please explain further with your prevention scheme.
__________________
The Butcher 

28th January 2007, 09:56 PM  #1576 
diyAudio Member
Join Date: Sep 2006

Hi Chris ma
sorry to hear of more troubles. Not sure definitely whether your problem was loose transistors. Screw fittings don't usually work loose unless they're vibrating, but maybe silicone washers "unstretched"? I wonder if you have oscillation causing the VAS transistor to reverse bias the upper output transistor.... it would suffer proportionally inreasing damage depending on the volume level in the amp. But an overtemperature limit would still be worht it I think! cheers John 
29th January 2007, 07:11 AM  #1577 
diyAudio Member
Join Date: Aug 2005

Really sory to read about those dead speakers. Thats why I still don't dare to go for it with my newly constructed JLH's (the are still running on Ig = 150 mA).
I'm now contructing the Rod elliot speaker protection circuit. That will give me some pease of mind... Anyone has any experience on this one? Glad to hear some comments. greetings, MArco 
29th January 2007, 07:12 AM  #1578 
diyAudio Member
Join Date: Aug 2005

Maths ...
Now that both my channels are playing (Iq of 0.15 A) it's about time to get into the maths (or should I have done that before hand ;)). Please help me with appropriate comments.
As I've acquired some really massive heatsinks there is no need to take these into account. The thermal resistance from transistor to sink will be the limiting factor thermal wise. Now we've that out of the way let's comtinue... The rails are 24 Volts (plus and minus that is); lets asume 3 V losses (as per Rod Elliots article). I expect the impedance of my speakers to stay above 5.5 Ohms. The Iq to take full advantage of the voltage swing to produce max power in 5.5 0hms then has to be: U = V/R = 21/ 5.5 = 3.8 = Iq min. The delivered power into 5.5 ohms will then be: P= UV = 3.8 * 21 = 80 Watts. This is incorrect as the ouput power should be RMS and this is calculated peak to peak. Should it then be P = 3.8 * SQRT(21) = 17 Watts ? or P = 3.8 * 2 * SQRT(21) = 35 W? My guestimate will be the second option. Is that correct? With no music playing there is no other place for the current to go than into heat in the heat sinks. So the heat to dissipate will be 24 * 3.8 = 91 W. This will be shared over the number of output resistors. So with two that will be 45W dissipation of heat per transistor. In the high power version with 4 transistors that will be 23 W per transistor. According to this reasoning the heat dissipation has to decrease when playing music. The energy used to drive the speaker is no longer avaiable to heath up the heat sink. Of course it is not wise to rely on this to downsize the neaded heat sinks. Lets recalculate for 4 and 8 ohms load to compare to other amps: 4 Ohms: 4 Ohms ask for a higher current. As we have an Iq of 3.8 no more current will be delivered. So the max power into 4 Ohms will be either 17W or 35W. To be able to drive 4 ohms speakers to max power we should have an Iq = 21/ 4 = 5.25 A. Total heath dissipation will be 126 W. High power version needed to share this over 4 transistors to give 31 W per transistor. 8 Ohms: U= V/R = 21/ 8 = 2.6 Amps asked for at max power. The Iq of 3.8 allows for this. P = 2.6 * 2 * SQRT(21) = 24 W. Heath dissipation stays as it is (91 W). To continue the calculations with Iq = 3.8: The load on the power supply will be the same 91W (the dissipated heath). As I have two channels on one PS that will be a load of 180 W. I use 2 capacitance multipliers (one for each channel). Say that each one consumes 10 W. Assume the bridge rectifier als consumes 10W. Total load on the transformers is then 210 W. To allow for some safety margin the transformers should than be about 400 VA. I have 320 VA avaiable which seams to allow for some margin. Will they cope? Then lets push things a little. Lets drive the 4 ohms speakers. Iq of 5.25. Total power for two channels 250 W + 20 W (cap. multiplier) + 10 W (bridge) = 280 W. Will 320 VA transformers cope or smoke? Any comments on the above very much appreciated. MArco 
29th January 2007, 09:58 AM  #1579  
diyAudio Member
Join Date: Apr 2001
Location: UK

Re: Maths ...
Quote:
Unfortunately there are some errors in your calculations and rather than comment on each one I think it would be easier for me to start from scratch. You have +/24V rails and, as you rightly say, account must be taken of losses so you can expect a peak output voltage of around 21 to 22V. Let's use the worst case of 21V. The minimum impedance of your speakers is 5.5R which will mean a peak output current of 3.8A. The JLH circuit will deliver a peak output current of about 1.35 times the quiescent current so the quiescent current should be 2.8A. The power delivered into the load is given by Vpeak x Ipeak / 2 (21 x 3.8 / 2) which equals 40W. The maximum dissipation in each output transistor is equal to the product of the supply rail voltage and the quiescent current, ie 24 x 2.8 or 67W. This is too much for a single transistor to handle (even with very large heatsinks) so a parallel pair of output transistors will be necessary. Each of the four output transistors will then dissipate 33.5W. With a 4R load, the maximum power output will be limited by the available peak output current which is 3.8A. The maximum power output will be Ip^2 x R / 2 which equals 29W. For an 8R load, the maximum power output is limited by the available peak output voltage which is 21V. The maximum power output will be Vp^2 / R / 2 which equals 27.5W. So far as transformer sizing is concerned, I normally recommend a VA rating that is twice the product of the total transformer secondary voltage and the total quiescent current draw. If you are using a capacitance multiplier rather than a regulator, for +/24V rails your transformer secondary voltage will need to be 22Vrms per winding. The total quiescent current draw will be 5.6A so the minimum transformer VA rating should be 44 x 5.6 x 2 = 493VA. A 500VA transformer will be suitable. Geoff 

29th January 2007, 10:15 AM  #1580 
diyAudio Member
Join Date: Oct 2003

Geoff,
It is my (limited) experience that a factor of 3 in terms of toroidal VA will help to minimise of tranformer humm. But this is also dependent on the size of the capacitor immediately after the rectifiers, and of course the transformer design itself. The maximum ripple current can be easily 10 times of Iq with a 10% duty cycle. IMHO, the price and size difference between a 500VA and a 800VA is really minimal, compared to the rest (heatsinks, housing, and what not). Patrick 
Thread Tools  Search this Thread 


Similar Threads  
Thread  Thread Starter  Forum  Replies  Last Post 
20 Watt Class A amplifier  JOE DIRT®  Solid State  8  31st January 2012 11:41 PM 
JLH 10 Watt class A amplifier  DomKo  Solid State  3  18th November 2003 06:32 AM 
New To Site?  Need Help? 