View Full Version : AMP3 kit
Here are some features of AMP3 and the kit:[list]
Small footprint, 25x75 mm and 22 mm high. Perfect for very compact amplifiers and small bi-amped speakers.
2x25W from a single 14.5V supply
2x15W at 0.1% THD+N into 4 ohms
Sleep/awake mode. In sleep mode the current is about 0.3mA.
Double weight copper, double sided PCB with print on both sides
The power chip is placed on the bottom side of the PCB. The PCB can be placed flat on the surface of a heat sink or you can use the amplifier housing as heat sink
Toroid inductor filters for low EMI
Overheat and overcurrent protection shuts down amp on error
Oversized onboard bulk capacitors 2x220uF reduce the ripple drawn from power supply.
Low noise thin film resistors
Space for onboard through hole input capacitors RM2.5
On board (onchip) +5V with 100uF bulk filter / bulk capacitor
Shipping from end of mars 2005.
A question is what wire to wind the toroids with. For AMP3 there will be T60-2 toroids in the kit. This is the biggest toroids that could fit on the PCB. Ideally, you should wind the toroids in one full layer without overlapping. The inductance should be 10uH and with the supplied toroid you need 40 turns. With 0.5 mm (24AWG) there is space for 41 turns on this toroid. This wire diameter is if anything oversized compared to what is used for other tripath amps. So 0.5 mm it is!
An even better alternative is to use multiple thinner wire. This decreases the skin effect at high frequency and lowers losses. Four 0.25 mm wires could be used. Or 16 strands of 0.12 mm wire. Or 100 strands of 0.05 mm wire which is probably ideal! If you buy this ready made it is called Litz wire. This is simply twisted multistrand wire with each strand individually insulated. If you make or buy Litz wire, make sure it is of the type that you can solder wothout cleaning the end of each strand.... Litz wire is quite expensive to buy. You need about 0.8 meter per toroid.
The toroids on the picture of the first AMP3 prototype boards are undersized T44 toroids and the wire is two 0.25 mm wires, which is really to low and will probably run a bit hot at full power.
For comparison the AMP1/TA2022 uses T68-2 toroids and 0.6 mm (22AWG) wire, the AMP2 uses T108-2 toroids and 1.2 mm wire.
I have written some a text about power supplies for AMP3
here (http://www.41hz.com/main.aspx?pageID=106). It covers the pros and cons of battery supplies, unregulated and regulated power supplies for AMP3
I have a Sonic Impact T-Amp and have asked a technician to put the PCB in a proper case with better connectors and a built-in 12V sealed lead-acid battery. I have retained the stock volume control+ON/OFF switch. I am enjoying this portable, 1-input integrated amp but it looks like your AMP3 could be a drop-in replacement. Any thoughts on this application?
Hi Jan, I'm assuming that 12 volts will operate the amp3 without any problems with dynamics? Thanks...
Hi. Yes 12V is quite enough for dynamics, as long as you are not driving the amp into clipping. For a given speaker impedance, you can only get a limited output for a given input voltage. Driving the amp inputs above a level where the output "should" be higher than the rail will mean outputs will clip.
Hi Dalsung. Sorry I did not respond before. I think the AMP3 would be very nice to use in a similar way that you describe. It is not a drop in replacement for the T-amp board in the sense they do not have the same size / shape. The AMP3 should be a better spec alternative to the T-amp and my prototype runs well of a 12V battery.
I just have one quick question, EVERYTHING to get this amp running is included in the kit, so you only have to assemble it right ?
I know housing, psu and speaker and inputs bushes arent included but everything to mount on the pcb is ?
Are instructions included as well ?
Included: PCB and all onboard components, including toroids and I will also include copper wire to wind them.
You will add the following:
- Housing and/or heat sink and screws to fit the PCB.
- Power supply. 12 to 14.5V recommended. Works well with a 12V sealed led battery.
- Fuse holder / fuse.
Depending on your application, you may want to add the following:
- Volume potentiometer. A 50 k ohm dual pot should work fine.
- If your board will be hooked up by "anyone" you may want to add a power supply polarity protection diode.
Yes, assembly instructions will be included.
There are no pictures of the final board yet, as the PCB in its final form has not arrived. It should be here within a few days.
Ooohh allright then I understand! cant wait to see it, and cant wait to build my own!
jan, can you give us the latest info on the shipping date of AMP3?
I just recieved the last component shipment and will start shipping as soon as I have verified that these and the final version of the PCB:s are OK. If so, shipments will start in the middle of this week.
Hi Jan,
when will you ship AMP3?
have a nice weekend
greets
------
jens
Hi,
sorry for impatience; it actually was excitement:D
Kits arrived today and look good, so does the documentation.
Will let you know how everything turns out and give you a comprehensive review at the latest on sunday.
thanks+greets
------
jens
I am interested in trying to wind the coils using 0.05 mm Litz wire, but I have one lingering doubt: since every strand of the Litz wire is individually insulated, doesn't the thickness of the insulation increase the overall diameter to the point where 100 strands have too great a total diameter to wind 40 turns? Thanks!
Christopher Witmer
Tokyo
Hi,
Litz wire is insulated by vanish that is quite thin. Normally, the impedance and losses with litz wire would be lower so you could use a slighty thinner "diameter" equivalent. And there is margin for a few extra turns on the supplied toroids. A low number of overlapping turns will not degrade performance. Eventual overlapping should preferably not be "last turns over the first turns" but "last on the second last". So I am afraid you will have to test as it is hard to give exact numbers.
Toroid core question --
I am looking at MPP (MolyPermalloy Powder) cores to substitute for the T60-2 iron powder cores in the AMP3 kit, and I see that for a given size, there is a range of choices with regard to the inductance rating of the core. What inductance rating should I be looking for? For example, at one website (http://www.trio-tw.com/TC/TC.pdf) I see cores of appropriate dimensions (*) that have AL Values of 19, 34.5, 41.5, 47, and 50. THANKS !
-- Chris
(*) Dimensions of T60-2 core are 15.2mm (OD) x 8.53mm (ID) x 5.94mm (HT)
Definition of AL Value (nH/N2):
"The inductance rating of a core in nanoHenries per turn squared (nH/N2) based on a peak flux density of 10 gauss (1 milliTesla) at a frequency of 10 kHz. An AL value of 40 would produce 400 H of inductance for 100 turns and 40mH for 1000 turns."
Hi,
You should look for the lowest permeability (U) type. Higher permeability will give higher losses and fewer winding turns for a given core size and given inductance. The lowest permeability I have seen for MPP is 27U, but 60U is more available. As a comparison, the type 2 iron powder cores supplied with out kits has a permeability of 10U, so you will need fewer turns for the MPP material. But MPP generally has lower losses than iron powder so a higher permeability should be OK with MPP. The AL value is a calculated from the cross section area of the toroid and the U value of the material. So in case the AL values are for the same geometry, simply go for the lowest AL value.
One issue is that the cores should not saturate even at the highest currents, as a saturated output filter could cause damage, before overcurrent limiting shuts down. However, with T60 or T54 size cores, I do not think it will be a problem.
I would suggest substituting the cores on one channel and see if one runs warmer than the other at high power.
Keep us posted on how this works for you.
MPP is a high quality (and high price material). It should have lower losses and better linearity than iron powder. On the down side is that it has a lower saturation limit than iron powder. According to Tripath, the iron powder cores do not contribute significantly to THD+N, compared to the rest of the circuit. In my opinion, MPP could still be worth a try.
Powered by vBulletin® Version 4.2.0 Copyright © 2013 vBulletin Solutions, Inc. All rights reserved.