Dear All,I have list THAT 1200 in my ebay store.
THAT 1200 work like a input transformer and is developed by Jensen Transformer. 🙂 That is why JRDG prefer it.
Developed by Bill Whitlock of Jensen Transformers, the patented InGenius input stage uses clever bootstrapping to raise its common-mode input impedance into the megohm range without the noise penalty from the obvious solution of using high-valued resistors. Like transformers, InGenius line receivers maintain their high CMRR over a wide range of source impedance imbalances ¡X even when fed from single-ended sources. But unlike transformers, these wide bandwidth solid state devices offer dc-coupling, low distortion, and transparent sound in a small package at reasonable cost.
THAT 1200 work like a input transformer and is developed by Jensen Transformer. 🙂 That is why JRDG prefer it.
Developed by Bill Whitlock of Jensen Transformers, the patented InGenius input stage uses clever bootstrapping to raise its common-mode input impedance into the megohm range without the noise penalty from the obvious solution of using high-valued resistors. Like transformers, InGenius line receivers maintain their high CMRR over a wide range of source impedance imbalances ¡X even when fed from single-ended sources. But unlike transformers, these wide bandwidth solid state devices offer dc-coupling, low distortion, and transparent sound in a small package at reasonable cost.
Possible good chassis for making an IcePower Amp?
I just ran into these on a surplus site. If it looks as good as the picture, it should make a nice easy to do chassis for mounting a couple of the ICEPOWER boards. I'm tempted to try one for a preamp, too, but I can't tell whether it would be possible to put knobs on the front.
/encl_2.htmlhttp://www.surplussales.com/Cab-RacksHardware/encl_2.html
I just ran into these on a surplus site. If it looks as good as the picture, it should make a nice easy to do chassis for mounting a couple of the ICEPOWER boards. I'm tempted to try one for a preamp, too, but I can't tell whether it would be possible to put knobs on the front.
/encl_2.htmlhttp://www.surplussales.com/Cab-RacksHardware/encl_2.html
I bought 7 of those enclosures 3-4 months ago and was planning on using one for this project. It's very nice.
dnewma04 --
So, how practical would it be to put controls on the front panel with knobs (and still have it look decent, I mean)?
So, how practical would it be to put controls on the front panel with knobs (and still have it look decent, I mean)?
Sorry, I should have read your last post more thoroughly. It wouldn't be difficult at all. The front piece can be remove with 4 screws. My plan was to replace it with a piece of thick aluminum or some nice wood mounted in the same 4 positions. The are a couple of openings in the front panel that could be utilized. I'll try to get some pics posted of the unit with the front plastic panel removed.
Here is what the inside of the surplus chassis looks like after 15 minutes of stipping. It's a very very nice little chassis.
The front panel attaches with 6 screws and would be very easily swapped.
The front panel attaches with 6 screws and would be very easily swapped.
dnewma04--
Thanks for the photos. I'll be following right in your footsteps on this one it looks like. I probably ought to order a second chassis, stuff like that can be hard to find. I wonder how hard it would be to sand off the lettering and labels off the back.
That section on the left looks like a great place to add some EMI filtering for the AC line.
Thanks for the photos. I'll be following right in your footsteps on this one it looks like. I probably ought to order a second chassis, stuff like that can be hard to find. I wonder how hard it would be to sand off the lettering and labels off the back.
That section on the left looks like a great place to add some EMI filtering for the AC line.
Design considerations regarding class II, safety ground, chassis ground, signal groun
I am on holidays at the moment so no ‘real’ work can be done, so even they send the boards I can’t work on them or connect them or so. But I gave this grounding thing a second thought.
As written in the manual the final amplifier has to build as a class II device for safety, meaning the ground pin in your mains connector should not be used, in fact you should use a mains plug with only 2 pins. Your final amplifier should be build according the safety issueds for class II equipment.
But your amplifier should have ground for your signal. Class II and grounding your signal to chassis works and is allowed. I included the grounding of a Hypex class d module, these are also build as Class II equipment and uses grounding signal to chassis as well.
If you take a look at the photos of the Jeff Rowland 102, the Cary CAA1 and the Belcanto M300 you can see these are all have ground to the chassis. The first two are using the special grounding place on the board with label W200 (they all seem to have the jumper installed), the Belcanto uses the ground of +47VDC jumper (J1) for ground to chassis (at least it looks this way), this point is directly connected to the grounding point W200.
The commercial Jeff Rowland Model 102 and the (Thailand build) Murano E-300S are using additional input circuitry. Jeff Rowland uses the THAT 1200 device (works even from asymmetric to symmetric with great common mode rejection and signal is treated like a real transformer in terms of isolation). Murano is using a Jensen JT-10KB-DPC for doing so, this including some extra opamps. I am not sure why they use these extra circuitry, the only I can think of is the following. The balanced input of the ASC200 board uses a standard balanced input circuit using a MC33079 (a quad opamp) and a circuit like this achieves ‘only’ 40 dB of common mode rejection, this because of the used resistors. But experienced shows that if this number is higher your resolution of the system, the micros information is also better, so soundstage and other details like cymbals and voices have more ‘air’, more authority. On some Danish website you can see that people (can’t read Danish so don’t know who they are or their considerations) modify the Murano units with a transformer like the Lundahl LL7902 or LL1545A. Myself I believe the LL1527XL is more suitable, this also because of the price. The LL7902 sells for 100 euros or so, the LL1527XL you can buy for half of it. I feel this additional component is far to expensive, I mean the board itself is sells for 70 US Dollars, so about 50 euros. Would be funny to have an amp that costs 50 euros itself and will use a 100 euros of cost input circuit.
I like the idea to use an additional circuitry. So am thinking of the use of the THAT1200 device, but before this I will try first it without (less is better is not always correct I experienced myself). I also believe B&O is not correct on connection signal – and ground together, a true balanced input should be used when an asymmetric (so unbalanced cinch cable is used) in a way that signal + is connected to pin 2 and ground of the unbalanced cable to ground. The input – is not used. I believe for this reason you should give the amplifier first test run with really cheap speakers connected, this hum / hiss problem could destroy your speakers. But if B&O writes ground to signal – that we should believe it, I mean if they can come up with some great amplifier module design, they should not know how to make an input circuitry. Or?
Wim
I am on holidays at the moment so no ‘real’ work can be done, so even they send the boards I can’t work on them or connect them or so. But I gave this grounding thing a second thought.
As written in the manual the final amplifier has to build as a class II device for safety, meaning the ground pin in your mains connector should not be used, in fact you should use a mains plug with only 2 pins. Your final amplifier should be build according the safety issueds for class II equipment.
But your amplifier should have ground for your signal. Class II and grounding your signal to chassis works and is allowed. I included the grounding of a Hypex class d module, these are also build as Class II equipment and uses grounding signal to chassis as well.
If you take a look at the photos of the Jeff Rowland 102, the Cary CAA1 and the Belcanto M300 you can see these are all have ground to the chassis. The first two are using the special grounding place on the board with label W200 (they all seem to have the jumper installed), the Belcanto uses the ground of +47VDC jumper (J1) for ground to chassis (at least it looks this way), this point is directly connected to the grounding point W200.
The commercial Jeff Rowland Model 102 and the (Thailand build) Murano E-300S are using additional input circuitry. Jeff Rowland uses the THAT 1200 device (works even from asymmetric to symmetric with great common mode rejection and signal is treated like a real transformer in terms of isolation). Murano is using a Jensen JT-10KB-DPC for doing so, this including some extra opamps. I am not sure why they use these extra circuitry, the only I can think of is the following. The balanced input of the ASC200 board uses a standard balanced input circuit using a MC33079 (a quad opamp) and a circuit like this achieves ‘only’ 40 dB of common mode rejection, this because of the used resistors. But experienced shows that if this number is higher your resolution of the system, the micros information is also better, so soundstage and other details like cymbals and voices have more ‘air’, more authority. On some Danish website you can see that people (can’t read Danish so don’t know who they are or their considerations) modify the Murano units with a transformer like the Lundahl LL7902 or LL1545A. Myself I believe the LL1527XL is more suitable, this also because of the price. The LL7902 sells for 100 euros or so, the LL1527XL you can buy for half of it. I feel this additional component is far to expensive, I mean the board itself is sells for 70 US Dollars, so about 50 euros. Would be funny to have an amp that costs 50 euros itself and will use a 100 euros of cost input circuit.
I like the idea to use an additional circuitry. So am thinking of the use of the THAT1200 device, but before this I will try first it without (less is better is not always correct I experienced myself). I also believe B&O is not correct on connection signal – and ground together, a true balanced input should be used when an asymmetric (so unbalanced cinch cable is used) in a way that signal + is connected to pin 2 and ground of the unbalanced cable to ground. The input – is not used. I believe for this reason you should give the amplifier first test run with really cheap speakers connected, this hum / hiss problem could destroy your speakers. But if B&O writes ground to signal – that we should believe it, I mean if they can come up with some great amplifier module design, they should not know how to make an input circuitry. Or?
Wim
attached document belongs to the 'My design considerations regarding class II, safety ground, chassis ground, signal ground.' reply.
Wim
Wim
Do you mean that the [-] input should just be left floating?? If the input circuit were to be using an actual transformer (one without center ground on input side) then doing that would definitely be wrong. There would effectively be no input signal unless the [-] were connected.
The correct way to wire that situation would be to connect the [ground] of the single ended preamp to the [-] of the amp input, and the [hot] of the preamp to the [+] of the amp input; and also (usually) chassis ground to chassis ground for safety, assuming equipment that requires safety ground. 200ASC apparently doesn't require safety ground, but I would think there should be some signal connection to chassis in some manner. What is shown on your Ucd diagram seems to make sense.
I've been playing with single ended inputs and here is what I've found. It does not matter if the signal ground is connected to the the negative input only, ground only, or both negative input and ground. The sound is the same.
If the source signal ground is connected to the board ground, and then the source signal connected to negative input, it will be half as loud as when connected to the positive input only.
I think I will end up driving my boards with an OPA1632 converting single ended to balanced.
If the source signal ground is connected to the board ground, and then the source signal connected to negative input, it will be half as loud as when connected to the positive input only.
I think I will end up driving my boards with an OPA1632 converting single ended to balanced.
That's because balanced inputs using the traditional opamp circuits aren't the same as a transformer would be (see the writeup on the THAT/Jensen circuit that was referenced earlier on this page). The impedances to ground are wrong. Both inputs have a low common mode impedance to ground, so an input into one with the other floating will still give a strong output. The input that is left floating effectively is coupled to ground, whereas in a real transformer based balanced input a floating input couples to the non-floating one.
For the simpler single-opamp version, gain will be different depending on which is driven and which is left floating, too, because opening one effects the feedback resistor ratio and the other doesn't. With a real transformer (or the THAT circuit), the floating input would just float along with the non-floating one, resulting in no significant differential input and hence no output -- the current that flows between the differential inputs is what determines output. So by leaving the [-] input floating you'd be counting on the circuit being a crappy pseudo-balanced setup, and get no common mode noise rejection -- that won't work with the real thing.
For the simpler single-opamp version, gain will be different depending on which is driven and which is left floating, too, because opening one effects the feedback resistor ratio and the other doesn't. With a real transformer (or the THAT circuit), the floating input would just float along with the non-floating one, resulting in no significant differential input and hence no output -- the current that flows between the differential inputs is what determines output. So by leaving the [-] input floating you'd be counting on the circuit being a crappy pseudo-balanced setup, and get no common mode noise rejection -- that won't work with the real thing.
- Status
- Not open for further replies.