Jeff Rowland Model 9ti clone?

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
That would be two bridged groups of four parallel ICs. Although the specs of 70 A continuous and 140 A peak do not sound like four LM3886 in parallel at all. Even eight in parallel couldn't make that. Neither the TDA7293 and TDA7294, which JR uses in some of his designs. Well, let's put that down to marketing.

Adapt TABLE 1 from AN-1192 and you get a recommended ±28 V rail to rail for 2 Ohm and ± 37 V rail to rail for everything above 4 Ohm. That is a 2 x 22 V transformer for the 2 Ohm version and a 2 x 30 V transformer for the standard version.

Transformer size would be around double what AN-1192 uses, that is something around 1500 VA.

Heatsink size will become an issue, if you really want to use that amount of power. Fans will be necessary.

And then there is that CNC-machined front... :cannotbe:
 
anybody want to try? I believe they use the same LM3886 chips total 8 chips per channel to achieve this..

any ideas? recommendations? how about supply voltage for this.. how much would be fine before rectification?

lets make this if anybody is interested...

That would be two bridged groups of four parallel ICs. Although the specs of 70 A continuous and 140 A peak do not sound like four LM3886 in parallel at all. Even eight in parallel couldn't make that. Neither the TDA7293 and TDA7294, which JR uses in some of his designs. Well, let's put that down to marketing.

Adapt TABLE 1 from AN-1192 and you get a recommended ±28 V rail to rail for 2 Ohm and ± 37 V rail to rail for everything above 4 Ohm. That is a 2 x 22 V transformer for the 2 Ohm version and a 2 x 30 V transformer for the standard version.

Transformer size would be around double what AN-1192 uses, that is something around 1500 VA.

Heatsink size will become an issue, if you really want to use that amount of power. Fans will be necessary.

And then there is that CNC-machined front... :cannotbe:


Would be very interested myself , has to be able to drive 1 ohm as the H20 amplfiers ...
 
Would be very interested myself , has to be able to drive 1 ohm as the H20 amplfiers ...

And how much power do you want or need to get into 1 Ohm? In the end it comes down to putting as many ICs in parallel as you need to achieve the peak current.

E. g. a non-bridged amplifier with the maximum recommended rail voltage of ±38 V gives you a peak output voltage of ~33 V. 33 V and 1 Ohm requires a peak current of 33 A, i. e. at least 5 LM3886s (6 TDA7293/7294) without taking impedance dips and reactance into account. In theory you get 33 V x 33 A / 2 = 544,5 W output power into 1 Ohm and need something like a 1000..1500 VA transformer. In practice each IC will "see" a 5 Ohm load and probably overheat, if you ask it to really deliver that amount of power. The impedance asks for at least 8 LM3886s to make them at least heatsinkable and to provide the current a realistic speaker load with nominal 1 Ohm needs.

Same estimates for a bridged amplifier. Rail voltage of ±38 V gives you a peak output voltage of ~66 V -> 66A -> at least 18 LM3886s (20 TDA7293/7294) to provide the current, at least 32 LM3886s to provide the heatsinking capability and current for real speakers. Theoretic result 2178 W into 1 Ohm and the need for a three-phase outlet for the 4000..6000 VA transformer it takes to get that.

On the other hand, maybe you only want the usual 20..40 W most Gaincloners use. Then you only need a peak output of 9 V and 9 A for 1 Ohm, which means ±11,5 V rails, 2 LM3886s to provide the current, an 80..120 VA transformer and a very small heatsink, if any at all.
 
Last edited:
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.