Discrete vs. Non-Discrete

[ruffrecords]

There is nothing like a 5531. The original (single) was TDA1034 from Mullard/Philips/Signetics, which became 5534 when licenced to other foundries. The sibling is the dual 5532. There was also an extinct 5539 (single in 14-pin DIL package).

Don't know how they do it nowadays but back then (1970s) licencing was done by a mask exchange agreement. This meant other foundries not only used the identical circuit but also the identical physical die layout of the original. For the end user this meant chips from different manufacturers would perform identically. The only differences were how well each foundry controlled its processes.

Cheers

Ian

 

[JohnRoberts]

Don't know how they do it nowadays but back then (1970s) licencing was done by a mask exchange agreement. This meant other foundries not only used the identical circuit but also the identical physical die layout of the original. For the end user this meant chips from different manufacturers would perform identically. The only differences were how well each foundry controlled its processes.

Cheers

Ian

I do not want to open this can of worms but here have been differences between the same chips run on different foundries. Things like 1/F noise are often caused by impurities in the raw silicon substrate, or process differences. There are other potential differences too but like I said I don't want to open that can of worms, as it will give the tweakers more to tweak 

In general the chips all behave alike enough to be used interchangeably and perform properly, and when they don't they are found out and fall out of use.  Large companies (like when I worked at Peavey) will generally approve a chip part number and a single manufacturer for a given part, while some common parts were multi sourced. We had a special house number 5532 that was selected for noise, and I later brought in a cheaper jelly bean 5532 to use in low gain sockets (like inputs and outputs where IC noise floor was not likely to be audible).  We used enough of them I could justify two part numbers.

JR 

 

[ruffrecords]

I do not want to open this can of worms but here have been differences between the same chips run on different foundries. Things like 1/F noise are often caused by impurities in the raw silicon substrate, or process differences. There are other potential differences too but like I said I don't want to open that can of worms, as it will give the tweakers more to tweak 

In general the chips all behave alike enough to be used interchangeably and perform properly, and when they don't they are found out and fall out of use.  Large companies (like when I worked at Peavey) will generally approve a chip part number and a single manufacturer for a given part, while some common parts were multi sourced. We had a special house number 5532 that was selected for noise, and I later brought in a cheaper jelly bean 5532 to use in low gain sockets (like inputs and outputs where IC noise floor was not likely to be audible).  We used enough of them I could justify two part numbers.

JR

Back in the early 80s I headed up the design team that created the UK's version of the Tandy CoCo called the Dragon 32 (because it was made by Mettoy in Wales). We used a 4066 quad switch to select joystick pot voltages to the AtoD converter.  Mettoy made 250,000 units in the first year so you can imagine how were were inundated with offers from semiconductor manufacturers and Mettoy sourced common parts like the 4066 from several sources. One batch caused very odd random value problems with the joystick readings. Turned out the 4066 was oscillating at VHF.

Cheers

Ian

 

[JohnRoberts]

Back in the early 80s I headed up the design team that created the UK's version of the Tandy CoCo called the Dragon 32 (because it was made by Mettoy in Wales). We used a 4066 quad switch to select joystick pot voltages to the AtoD converter.  Mettoy made 250,000 units in the first year so you can imagine how were were inundated with offers from semiconductor manufacturers and Mettoy sourced common parts like the 4066 from several sources. One batch caused very odd random value problems with the joystick readings. Turned out the 4066 was oscillating at VHF.

Cheers

Ian

Yup nothing like large scale manufacturing to reveal any snakes lurking in the woodpile. 

I can imagine variation in the on resistance of CMOS transfer gates (like 4066) between different foundries but oscillation seems unexpected,  but that is why there is such a huge difference between making one of anything, and 100k...  :

JR

 

[ruffrecords]

Yup nothing like large scale manufacturing to reveal any snakes lurking in the woodpile. 

I can imagine variation in the on resistance of CMOS transfer gates (like 4066) between different foundries but oscillation seems unexpected,  but that is why there is such a huge difference between making one of anything, and 100k...  :

JR

You bet your bippy there is.

Cheers

ian

 

[JohnRoberts]

You bet your bippy there is.

Cheers

ian

been dere done dat...

JR

 

[clintrubber]

There is nothing like a 5531. The original (single) was TDA1034 from Mullard/Philips/Signetics, which became 5534 when licenced to other foundries. The sibling is the dual 5532. There was also an extinct 5539 (single in 14-pin DIL package).

FWIW, there's also the 5533, which is actually the 'true' dual version of the 5534 (unlike the 5532).
Hence the 5533 comes in a DIL16, to accomodate the additional pinning the 5534 has over the 5532.
I ran into a few in a parts-box.

Also FWIW, hadn't heard of the 5539 before, but it looks to be a different design/type - the datasheets show quite different topologies.

Best regards

 

[L´Andratté]

Thanks for all the lessons, everyone.

+1 

 

[Spino]

Yes and no. The 5534 will drive a 600 ohm load so you can use a 1:1 transformer for a balanced output.  Neve and API amps use a step up transformer so they increase their maximum output level. This means for a 1:2 step up they need to be able to drive a 150 ohm load which the 5534 cannot do. So with +-15V rails you will get +22dBu into a 600 ohm load with  5534. With a Neve or API you can get +26dBU. If you have a discrete op amp that will take +-24V rails then it could drive also +26dBu via a 1:1 transformer.

Some discrete op amps can drive very low loads. I think the Neve BA440 can drive a load as low as 4 ohms. It was sometimes used to drive a cue speaker in some broadcast console designs. I think it was capable of about 4 watts.

Cheers

Ian

Ian, thanks for explaining.

Just for the "explain like I'm 5", the max load an opamp can drive is given by its V/mA capabilities? In the first 5534 datasheet I went reading there's no such thing as a "minimum RLoad '".

I guess I'm still at Ohm's law.

 

[abbey road d enfer]

Just for the "explain like I'm 5", the max load an opamp can drive is given by its V/mA capabilities? In the first 5534 datasheet I went reading there's no such thing as a "minimum RLoad '".

There's this: "Peak-to-Peak Output Voltage Swing 32 V Typ With VCC± = ±18 V and RL = 600 Ω", and that's pretty much the maximum output that the 5534 is capable of.
The ONSEMI datasheet has some more info; see Output swing on page 3.
https://www.onsemi.com/pub/Collateral/NE5534-D.PDF

 

[ruffrecords]

Ian, thanks for explaining.

Just for the "explain like I'm 5", the max load an opamp can drive is given by its V/mA capabilities? In the first 5534 datasheet I went reading there's no such thing as a "minimum RLoad '".

I guess I'm still at Ohm's law.

As Abbey has explained, the maximum (power) output is usually obtained with a 600 ohm load which is the lowest load the chip is designed to work into and this power works out at around 140mW. Any more than this and either the transistors will be operating outside their design parameters or, more likely, the chip dissipation will be exceeded, and the dvice temperature will increase until it burns out.  However, this may not be the maximum load it can drive. It may well be able to drive a 300 ohm load for instance but at a much lower voltage swing. The open loop gain will also suffer. The data sheet quotes the large signal open loop gain into 600 ohms as about 4.4dB less than it is into a 2K load. It will be even less than that into a 300 ohm load.

Cheers

Ian