2024-01-16
The OSCONIQ 3737 does not play a major role in the flashlight industry. This LED is used in only a few lights, most as a replacement for Cree XP-G3 or Nichia emitters.
The PUSTA1.PM emitters tested here come from a Lumintop Ring King, a light powered by 14500, which is specified by the manufacturer with a maximum of 1000 lumens. BLF user 1stein provided me with these emitters while swapping them, thank you very much!
Tj 85 °C, If 1,050 mA
- Type: single die, lateral chip
- Bin: unknown
- Color group: unknown
- CRI: min. 70, typ. 72 (R9 min. -50)
- Rated voltage: min. 2.60, typ. 2.75, max. 3.10 V
- Max. Forward current: 3,000 mA
- Max. Peak current: 4,000 mA
- Viewing angle: 120 °
- Thermal resistance: 1.9 K/W
- Max. Temperature Tj: max. 150 °C
Datasheet can be downloaded here: Datasheet (newest version, Osram)
At first glance, the PUSTA1.PM looks like an LED in the mid-power class. The substrate is made of white heat-resistant epoxy or plastic, with orange-brown areas visible on the sides, possibly from the copper or ceramic core, which ensures heat dissipation and contacting.
Viewed from above, the LED could be reminiscent of the old XP-G2, especially due to the two bondings on two outer corners of the LED chip. However, it is not only the phosphor that is covered here, but also the border under the entire dome. This usually has an effect on the color homogeneity in the light pattern (beam) with reflectors or lenses, often referred to as “cree rainbow”.
The PUSTA1.PM is designed in 3737 format, i.e. it is 3.7 x 3.7 mm in size.
The footprint does not correspond to the 3535 format. Nevertheless, the use of 3535 or XP boards is possible without any problems. During the reflow process, it is important to ensure that the thermal pad and contacts are aligned evenly.
This places the 3737 footprint fairly centrally between the 3535 and 4040 footprint. The latter was used by Lumileds in the Luxeon V CSP LED from 2017, an emitter that never became established in flashlights.
Due to the unusual footprint, accessories are hardly available; however, the symmetrical design means that rotated centering aids can be used.
The thermal pad is not electrically insulated, but connected to the cathode as with earlier Osram LEDs. This makes the use of DTP boards much more difficult, as the LED is electrically insulated from the body of the flashlight to prevent operation bypassing the driver electronics (“direct drive”). Depending on the voltage level of the battery, this usually leads to the destruction or massive reduction in the lifespan of the LED.
The LES is similar to the XP-G2, but the yellow ring around the illuminated area produces different colors: classic blue above the LED chip (as is to be expected from the CCT given in the order code), but yellow around the illuminated area. This phenomenon is also known from the XP-L2 or XP-G3.
The LED chip appears to emit light from the side, as is already the case with many modern emitters in flip-chip design. This would be interesting in that this LED would be one of the few that would have such a radiation characteristic in the classic lateral design. The older LE UW Q8WP - also from Osram - has a similar feature.
The luminous area is 6.9 mm² in size. Due to sideways emitted light it is difficult to determine the real LES size by only measuríng it.
The LED chip is connected using two bonding wires. Compared to modern emitters in flip chip design (which are contacted from below), the disadvantage here is the increasing forward voltage with increasing operating current. This could limit the maximum performance or even destroy the bonding wires if the current is too high, before the maximum luminous flux has even been reached.
Within official parameters, as far as known:
- at 3,000 mA (official maximum current): 1318 lm @ 3.07 V
- Power at official maximum: 9.21 W
- Efficiency at 3,000 mA: 143 lm/W
The OSCONIQ 3737 PUSTA1.PM tested here is very efficient. Although Osram generally only specifies wide binnings - for luminous flux as well as voltage and tint - the sample tested here is one of the most efficient 3V LEDs I ever tested. Above all, the low Vf of 3.07 V at the official maximum current means good constant current regulation even with classic 105c drivers, as the voltage only drops below this level when the battery is almost empty.
Overcurrent:
- Maximum reached at 7.2 A, at this point 2473 lm @ 3.54 V
- Power at maximum 25.5 W
- Sweet spot at about 5 A (1965 lm @ 3.27 V)
- Power at sweet spot 16.4 W
- Efficiency at maximum 97 lm/W
- Efficiency in the sweet spot 120.2 lm/W
NOTE: The test was aborted at 7.2 A to avoid destroying the emitter! In the case of emitters with bonding wires, this is usually indicated by a sudden significant increase in Vf as the operating current increases.
In general, the PUSTA1.PM offers high performance. The low specified thermal resistance definitely comes into play here and shows once again how important a low thermal resistance is for the discipline of maximum performance. From around 5 to 5.5 A, the Vf starts to increase, presumably due to the increasing resistance of the bonding wires caused by the positive temperature coefficient (PTC thermistor). In the operating range up to 5 A, there should be no problems in the longer term, assuming good cooling.
If the bonding wires were bigger in dimension, the maximum possible operating current would be around 9-10 A with presumably around 2800-3000 lumens. Ultimately, the area of the thermal pad would primarily limit the maximum possible current, as was the case with the Luxeon MZ, for example.
A comparison with other emitters in the similar power class is interesting. The impressively low Vf is all the more evident here, especially for an emitter with classic bonding. Even the 519A has a higher Vf. The efficiency is also impressive. Despite its smaller illuminated area, it easily achieves the efficiency of an XM-L2 U2 (determined as Bin U4 in the test), with a massively reduced Vf. In general, the PUSTA1.PM tested here is more in the performance and efficiency class of 5050 emitters with a correspondingly larger illuminated area.
Values at 25 °C Tsp, at 85 °C Tsp values are 13 % lower
The luminance is low. Due to the large luminous surface and the lateral radiation from the LED chip, the luminance drops further and is then in the range of conventional emitters such as XM-L2 1st gen, XP-L2, Nichia 519A óor Lumileds HL2X. This emitter is therefore only suitable for throwers to a limited extent.
The light image with reflectors is as you would expect: yellow rings line the blue spot. Nevertheless, this looks worse in the picture than it actually is, and is also less intense than with XP-L2, XP-G3 and the like. With suitable reflectors, this effect should be kept within limits.
The PUSTA1.PM tested here is classified by the manufacturer in the 70 CRI range. The spectrum shows no surprises; the color rendering is also within the expected range and the manufacturer’s specification of typ. 72 CRI is met. The duv is very good; there are no disturbing color casts when the light is completely mixed. At over 7500 K, the sample shown here is one of the colder specimens.
- Ra: 73
- R9: -20
- CCT: 7692 K
- duv: -0.0007
The OSCONIQ 3737 PUSTA1.PM is a very efficient LED. The Vf is also very low and, despite the smaller footprint, it can easily keep up with XM-L2 and other 5050 LEDs. I don’t like the classic lateral design, as there is a risk of the bonding wires being destroyed at more than 7 A. The light image is OK, but doesn’t come close to 519A and the like. The biggest disadvantage here, however, is the thermal pad, which is not electrically insulated; this makes it much more difficult to use this LED in the high-current range.
For normal flashlight applications where a DTP board is not absolutely necessary and high color rendering and a perfect light image are also not important, the PUSTA1.PM is suitable without any problems.
Pro
- very efficient
- very low Vf
- good tint (sample tested here)
- variants with high color rendition available (PUSTA1.CM)
Neutral
- beam quality in reflectors could be better
Contra
- limited overcurrenting due to bonding wires
- no electrically insulated thermal pad
- unusual footprint (3737)
Thank you for reading the test. 
Very promising efficieny wise combined with low duv 
