New TK2050 board

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Alternative Power Supplies?

S-350-24

Quote:
Originally Posted by cragthehak
I'm about to order one of these boards. From the sounds of things, would a smaller supply like this one at 24v 6A work better for this?

145W 24V DC 6A Regulated Switching Power Supply [K018] on eBay.ca (item 200475424389 end time 22-Jun-10 03:24:46 EDT)


Unless size is an issue, I would spend a little more and get the S-350-24 or 27v MeanWell power supply.
__________________
Scott

This thread is getting really interesting with input from designer and knowledgeable members. My contribution is merely to point to an alternative method of obtaining a decently charged +24Vdc SMPS at a much lesser cost. It's on Decibel Dungeon detailing the connection of 2x computer SMPS's. http://http://myweb.tiscali.co.uk/nuukspot/decdun/gainclonesmps.html
Although it's really for +/- supply there are some details on how to obtain +24V supply. I haven't built such a supply but would like to know the reactions and experience of members as to the technical diy aspects of such a project.

Thanks to all contributors, particularly Nick and hifimediy!
Best regards.
Joe A
 
What's wrong with my linking a website. It already happened to me before. I click the Insert Link icon, paste a copied version of the site on window and select OK.

http://http://myweb.tiscali.co.uk/nuukspot/decdun/gainclonesmps.html

Hi, you have "http://" doubled in the link. It's the reason why the link doesn't work.

BTW, I have ordered this unbelievably cheap SMPS for the TK2050 amp board. I haven't received it yet so I am curious how it will work.
Wholesale 150W SMPS Product-150W +5V/1A/+ 15V/1A/24V/5A ATX Switch Power Supply
 
Shouldn't we look for SMPS with as high switching frequency as possible?
The S-350-24 or 27v Mean-Well's has 25kHz switching frequency - but I believe there are models with >40kHz -like the above item - shouldn't they be better or would 25kHz is good enough with these kind of digital amp IC's ?
Gal
 
Use the cheap SMPS only with caution:
It needs load on all other rails in order to function properly (no smoke).
Connecting more than 1 12v supply in series might work, but the cheap meanwell 24v supplies are designed for 24v, and are even cheaper.
Further, they are smaller.

If you have the money to invest in this 2050 board, then you have money to buy a decent smps to go with it.
If you dont, dont buy any board, or buy the sure board, which is much more moddable.
*goes soldering his tk2050 sure board once more :p
 
Right. we have to choose another power stage IC for 200W use.

My idea was to use a 12V supply (for portable application 1 battery is lighter than 2 batteries :))

I'm referring to http://www.st.com/stonline/products/literature/an/10606.pdf in my comments below:

I suppose the amp uses two STA505 output stage (one per channel) in parallel BLT configuration (Figure 7 of spec)

But according to figure 11, you can parallel the two ST505 in parallel BLT for higher current capability.

Since I plan on using 12V, I won't be reaching anywhere close to 100W. maybe 25W? as other lower voltage implementation of tripath chips.

Thus, is it possible for one to modify the output and get more power out of a 12V battery? It seems straightforward enough, connecting the output of L1 with L3 and the output of L2 with L4
 
Thank you for your advices, teamacc. Actually, I wanted to order Meanwell 27V SMPS originally but it's a bit bulky and I wanted my new SMPS to fit into particular metal case. The one mentioned above seemed to be good in this aspect and its low price is only a nice bonus. :)

My understanding is that I do not have to connect two 12V rails in series, it seems to have one 24V, plus +-15V and +5V.

Regarding the price, I paid $25 incl. shipping and custom modification of main output voltage to 36V (well, there were also some other items in my order). I haven't found so cheap Meanwell SMPS anywere yet.
 
Repeat post with live link.

Thanks TCx

This thread is getting really interesting with input from designer and knowledgeable members. My contribution is merely to point to an alternative method of obtaining a decently charged +24Vdc SMPS at a much lesser cost. It's on Decibel Dungeon detailing the connection of 2x computer SMPS's. SMPS for the Gainclone chip amp

Although it's really for +/- supply there are some details on how to obtain +24V supply. I haven't built such a supply but would like to know the reactions and experience of members as to the technical diy aspects of such a project.

Thanks to all contributors, particularly Nick and hifimediy!
Best regards.
Joe A
 
This thread is getting really interesting with input from designer and knowledgeable members. My contribution is merely to point to an alternative method of obtaining a decently charged +24Vdc SMPS at a much lesser cost. It's on Decibel Dungeon detailing the connection of 2x computer SMPS's. http://http://myweb.tiscali.co.uk/nuukspot/decdun/gainclonesmps.html
Although it's really for +/- supply there are some details on how to obtain +24V supply. I haven't built such a supply but would like to know the reactions and experience of members as to the technical diy aspects of such a project.
I explored the same idea in the past, realizing that it's dangerous, because if the 'response time' of the switching controller on the first SMPS is not *identical* to the 'response time' on the other SMPS, they could start to oscillate... damaging the TP2050 board, themselves or more simply stop working (entering in protection mode).
So I don't suggest you to try that configuration, 2 switching controllers into the same PSU are never a good idea, although I have no doubt that in most cases it works.
Bye!
 
Thanks for all suggestions. here are new ways to tight the heatsink.
space.php
You added the white spacers to the bolts that retain the heat sink, I'm correct?
Thanks, bye!!
 
I explored the same idea in the past, realizing that it's dangerous, because if the 'response time' of the switching controller on the first SMPS is not *identical* to the 'response time' on the other SMPS, they could start to oscillate... damaging the TP2050 board, themselves or more simply stop working (entering in protection mode).
So I don't suggest you to try that configuration, 2 switching controllers into the same PSU are never a good idea, although I have no doubt that in most cases it works.
Bye!
Also please notice that a typical AT/ATX PC SMPS performs the feedback on the +5V line (more is the current consumption on that line, more the switching controller widens the pwm to mantain the correct +5V voltage).
A typical move to have a good +12V output is to apply a fixed load on the +5v line:
"One note of caution - most power supplies require a minimum load on the main output (5V in your case) - indeed some PSUs fail to start up (or even blow up!) without one. It seems like you've been ok so far, but (perhaps surprisingly) it would improve long-term reliability to draw maybe half an amp from the 5 volt output - perhaps using a 10 ohm, 5W wirewound resistor. Doing this should improve the regulation of the other outputs, and might alter the sound quality..."

...but in such way you got a fixed regulation of the +12v output, so more is the current consumption on the +12v line, more is the voltage drop on that line... because the pwm is fixed, based only on the +5v resistor current absorption.
To solve that issue you have to modify the feedback circuit so that it reads the voltage drop on the +12v line instead of the +5v line.
Do that and you will have a true +12v regulated output without the necessity of applying a resistor on the +5v line.
I did that modification on some old AT PSU and they worked fine.
Of course that modification requires knowledge of how a SMPS works.
Basically you have to connect the feedback line to the +12v output instead of the +5v, decreasing properly the +12v with a zener diode so that the switching controller "believes" to regulate the +5v but in reality is regulating the +12v line.
Choosing the proper zener diode is also possible to achieve over +15v on the +12v output, so for example you could connect 2 x 14.5v modified PSU obtaining a 29v output... just perfect for the TK2050.

I prefer a single SMPS for the reasons I explained above but I think that I could try with a couple of SMPS connected in series... I have 5 x 300W ATX identical power supply at home that are waiting to be used in any way!
Bye!
 
...

I'm referring to http://www.st.com/stonline/products/literature/an/10606.pdf in my comments below:

I suppose the amp uses two STA505 output stage (one per channel) in parallel BLT configuration (Figure 7 of spec)

But according to figure 11, you can parallel the two ST505 in parallel BLT for higher current capability.

...

Now that I think of it, wouldn't paralleling the two ST505 support a mono configuration at 200W? This would be nice for a sub amp. It seems (from reading some of the posts in the sureelectronics thread) that the issue is to null the outputs to zero DC.
 
Last edited:
Now that I think of it, wouldn't paralleling the two ST505 support a mono configuration at 200W? This would be nice for a sub amp. It seems (from reading some of the posts in the sureelectronics thread) that the issue is to null the outputs to zero DC.

Further reading into the spec:


"To drive loads having very low impedance obtaining higher output power it is possible to furtherly
connect together on the same load two STA50x IC's (each one already configured as
single-paralleled bridge) doubling the current capability.
The correct connection is putting together the inductor (L1 with L3 and L2 with L4) terminals
on the load side. This to avoid uncontrolled shootrough current spikes, due to time mismatch
between the two IC's, which could be possible if the outputs are directly paralleled on the same
inductor.
The L and C values of the output filter have to be calculated tacking in account the load impedance:
to be noted that the inductance is now provided by two self-inductors in parallel (L1/
/L3 and L2//L4), so each one must have double value.
The inputs must be driven in parallel."

I guess some modification in the output filter is required... But certainly a mono configuration with two sta505 in parallel is possible (idea for next version)
 
STA517 still not available

Most alike yes, not the best. In all aspects the STA516/517 are the best replacement, but probably cost a bit more...

Or some Apogee chips also are capable of more current/higher voltage, but I never tried those...
It is strange that the STA517 or 516 is still not available in the US.
.
STMicroelectronics | Digital Audio Amplifiers
.
I can buy STA508, 505, or 518 though.
 
Most alike yes, not the best. In all aspects the STA516/517 are the best replacement, but probably cost a bit more...
With STA517 you have to double the voltage input... and that becomes another project (different power supply and some adjustments on the board).
Could be interesting, but instead of being an improvement I think that it would be something different.
Bye!
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.